{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.034312390585489,
  "eval_steps": 3000,
  "global_step": 39000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742087173461915,
      "epoch": 0.00038902937171756465,
      "grad_norm": 5.03125,
      "learning_rate": 2e-06,
      "loss": 10.7355,
      "mean_token_accuracy": 0.0,
      "num_tokens": 8095.0,
      "step": 5
    },
    {
      "entropy": 10.742138862609863,
      "epoch": 0.0007780587434351293,
      "grad_norm": 4.90625,
      "learning_rate": 4.5e-06,
      "loss": 10.7472,
      "mean_token_accuracy": 0.0,
      "num_tokens": 17329.0,
      "step": 10
    },
    {
      "entropy": 10.742084884643555,
      "epoch": 0.001167088115152694,
      "grad_norm": 5.1875,
      "learning_rate": 7e-06,
      "loss": 10.7161,
      "mean_token_accuracy": 0.0,
      "num_tokens": 25729.0,
      "step": 15
    },
    {
      "entropy": 10.74206018447876,
      "epoch": 0.0015561174868702586,
      "grad_norm": 5.03125,
      "learning_rate": 9.5e-06,
      "loss": 10.658,
      "mean_token_accuracy": 0.00011547344038262964,
      "num_tokens": 33980.0,
      "step": 20
    },
    {
      "entropy": 10.741970634460449,
      "epoch": 0.0019451468585878235,
      "grad_norm": 4.71875,
      "learning_rate": 1.2e-05,
      "loss": 10.5514,
      "mean_token_accuracy": 0.0035519652417860927,
      "num_tokens": 41675.0,
      "step": 25
    },
    {
      "entropy": 10.741633415222168,
      "epoch": 0.002334176230305388,
      "grad_norm": 4.1875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.4556,
      "mean_token_accuracy": 0.030741829052567483,
      "num_tokens": 50277.0,
      "step": 30
    },
    {
      "entropy": 10.740493106842042,
      "epoch": 0.0027232056020229526,
      "grad_norm": 3.5625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3206,
      "mean_token_accuracy": 0.042657143995165826,
      "num_tokens": 58790.0,
      "step": 35
    },
    {
      "entropy": 10.737204456329346,
      "epoch": 0.0031122349737405172,
      "grad_norm": 2.984375,
      "learning_rate": 1.95e-05,
      "loss": 10.1918,
      "mean_token_accuracy": 0.041701832041144374,
      "num_tokens": 67247.0,
      "step": 40
    },
    {
      "entropy": 10.730194187164306,
      "epoch": 0.003501264345458082,
      "grad_norm": 2.515625,
      "learning_rate": 2.2e-05,
      "loss": 10.0582,
      "mean_token_accuracy": 0.04363730065524578,
      "num_tokens": 75174.0,
      "step": 45
    },
    {
      "entropy": 10.722168254852296,
      "epoch": 0.003890293717175647,
      "grad_norm": 2.203125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.969,
      "mean_token_accuracy": 0.04705791398882866,
      "num_tokens": 83203.0,
      "step": 50
    },
    {
      "entropy": 10.716745758056641,
      "epoch": 0.004279323088893211,
      "grad_norm": 2.03125,
      "learning_rate": 2.7e-05,
      "loss": 9.9154,
      "mean_token_accuracy": 0.05126173123717308,
      "num_tokens": 91798.0,
      "step": 55
    },
    {
      "entropy": 10.712809944152832,
      "epoch": 0.004668352460610776,
      "grad_norm": 1.9375,
      "learning_rate": 2.95e-05,
      "loss": 9.91,
      "mean_token_accuracy": 0.04944804683327675,
      "num_tokens": 100869.0,
      "step": 60
    },
    {
      "entropy": 10.710667610168457,
      "epoch": 0.0050573818323283405,
      "grad_norm": 2.03125,
      "learning_rate": 3.2e-05,
      "loss": 9.8383,
      "mean_token_accuracy": 0.0466838538646698,
      "num_tokens": 109697.0,
      "step": 65
    },
    {
      "entropy": 10.707512664794923,
      "epoch": 0.005446411204045905,
      "grad_norm": 2.046875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.7779,
      "mean_token_accuracy": 0.04920548666268587,
      "num_tokens": 118656.0,
      "step": 70
    },
    {
      "entropy": 10.699565601348876,
      "epoch": 0.00583544057576347,
      "grad_norm": 1.78125,
      "learning_rate": 3.7e-05,
      "loss": 9.758,
      "mean_token_accuracy": 0.047676689177751544,
      "num_tokens": 127748.0,
      "step": 75
    },
    {
      "entropy": 10.6899094581604,
      "epoch": 0.0062244699474810344,
      "grad_norm": 1.875,
      "learning_rate": 3.95e-05,
      "loss": 9.6559,
      "mean_token_accuracy": 0.052025371417403224,
      "num_tokens": 136621.0,
      "step": 80
    },
    {
      "entropy": 10.678217601776122,
      "epoch": 0.006613499319198599,
      "grad_norm": 1.953125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.5282,
      "mean_token_accuracy": 0.06281816884875298,
      "num_tokens": 145793.0,
      "step": 85
    },
    {
      "entropy": 10.661566066741944,
      "epoch": 0.007002528690916164,
      "grad_norm": 1.8515625,
      "learning_rate": 4.45e-05,
      "loss": 9.4566,
      "mean_token_accuracy": 0.06447997651994228,
      "num_tokens": 154466.0,
      "step": 90
    },
    {
      "entropy": 10.639564514160156,
      "epoch": 0.007391558062633729,
      "grad_norm": 1.921875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4185,
      "mean_token_accuracy": 0.06300977356731892,
      "num_tokens": 163297.0,
      "step": 95
    },
    {
      "entropy": 10.602224636077882,
      "epoch": 0.007780587434351294,
      "grad_norm": 1.765625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.2964,
      "mean_token_accuracy": 0.066444294154644,
      "num_tokens": 171747.0,
      "step": 100
    },
    {
      "entropy": 10.547998046875,
      "epoch": 0.008169616806068859,
      "grad_norm": 1.6171875,
      "learning_rate": 5.2e-05,
      "loss": 9.2224,
      "mean_token_accuracy": 0.061835135146975514,
      "num_tokens": 180499.0,
      "step": 105
    },
    {
      "entropy": 10.506940460205078,
      "epoch": 0.008558646177786422,
      "grad_norm": 1.71875,
      "learning_rate": 5.45e-05,
      "loss": 9.1182,
      "mean_token_accuracy": 0.06274112537503243,
      "num_tokens": 189227.0,
      "step": 110
    },
    {
      "entropy": 10.470497703552246,
      "epoch": 0.008947675549503988,
      "grad_norm": 1.6171875,
      "learning_rate": 5.7e-05,
      "loss": 8.9987,
      "mean_token_accuracy": 0.06519487500190735,
      "num_tokens": 198388.0,
      "step": 115
    },
    {
      "entropy": 10.417663478851319,
      "epoch": 0.009336704921221552,
      "grad_norm": 1.5,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.8933,
      "mean_token_accuracy": 0.0642558429390192,
      "num_tokens": 206854.0,
      "step": 120
    },
    {
      "entropy": 10.3519455909729,
      "epoch": 0.009725734292939117,
      "grad_norm": 1.546875,
      "learning_rate": 6.2e-05,
      "loss": 8.7861,
      "mean_token_accuracy": 0.0641202162951231,
      "num_tokens": 215506.0,
      "step": 125
    },
    {
      "entropy": 10.284876441955566,
      "epoch": 0.010114763664656681,
      "grad_norm": 1.4140625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6622,
      "mean_token_accuracy": 0.06737961098551751,
      "num_tokens": 223717.0,
      "step": 130
    },
    {
      "entropy": 10.18412799835205,
      "epoch": 0.010503793036374246,
      "grad_norm": 1.3046875,
      "learning_rate": 6.7e-05,
      "loss": 8.5306,
      "mean_token_accuracy": 0.067349299415946,
      "num_tokens": 232480.0,
      "step": 135
    },
    {
      "entropy": 10.091043281555176,
      "epoch": 0.01089282240809181,
      "grad_norm": 1.46875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.4423,
      "mean_token_accuracy": 0.06387559846043586,
      "num_tokens": 241350.0,
      "step": 140
    },
    {
      "entropy": 9.962540531158448,
      "epoch": 0.011281851779809376,
      "grad_norm": 1.34375,
      "learning_rate": 7.2e-05,
      "loss": 8.3445,
      "mean_token_accuracy": 0.06567118875682354,
      "num_tokens": 250610.0,
      "step": 145
    },
    {
      "entropy": 9.842002868652344,
      "epoch": 0.01167088115152694,
      "grad_norm": 1.1640625,
      "learning_rate": 7.45e-05,
      "loss": 8.1965,
      "mean_token_accuracy": 0.06004151217639446,
      "num_tokens": 259672.0,
      "step": 150
    },
    {
      "entropy": 9.687075901031495,
      "epoch": 0.012059910523244505,
      "grad_norm": 1.1796875,
      "learning_rate": 7.7e-05,
      "loss": 8.1149,
      "mean_token_accuracy": 0.06366611644625664,
      "num_tokens": 267826.0,
      "step": 155
    },
    {
      "entropy": 9.486919116973876,
      "epoch": 0.012448939894962069,
      "grad_norm": 1.0703125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.0603,
      "mean_token_accuracy": 0.06088076755404472,
      "num_tokens": 277324.0,
      "step": 160
    },
    {
      "entropy": 9.344527530670167,
      "epoch": 0.012837969266679634,
      "grad_norm": 0.95703125,
      "learning_rate": 8.2e-05,
      "loss": 7.8628,
      "mean_token_accuracy": 0.06783541738986969,
      "num_tokens": 285482.0,
      "step": 165
    },
    {
      "entropy": 9.084899616241454,
      "epoch": 0.013226998638397198,
      "grad_norm": 0.9140625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.8199,
      "mean_token_accuracy": 0.06509357057511807,
      "num_tokens": 293684.0,
      "step": 170
    },
    {
      "entropy": 8.910903358459473,
      "epoch": 0.013616028010114764,
      "grad_norm": 0.83203125,
      "learning_rate": 8.7e-05,
      "loss": 7.7429,
      "mean_token_accuracy": 0.06691662222146988,
      "num_tokens": 302605.0,
      "step": 175
    },
    {
      "entropy": 8.674479103088379,
      "epoch": 0.014005057381832327,
      "grad_norm": 0.84375,
      "learning_rate": 8.95e-05,
      "loss": 7.7506,
      "mean_token_accuracy": 0.06571518406271934,
      "num_tokens": 310849.0,
      "step": 180
    },
    {
      "entropy": 8.626084232330323,
      "epoch": 0.014394086753549893,
      "grad_norm": 0.78515625,
      "learning_rate": 9.2e-05,
      "loss": 7.7323,
      "mean_token_accuracy": 0.06051843762397766,
      "num_tokens": 319752.0,
      "step": 185
    },
    {
      "entropy": 8.481939506530761,
      "epoch": 0.014783116125267459,
      "grad_norm": 1.0,
      "learning_rate": 9.45e-05,
      "loss": 7.749,
      "mean_token_accuracy": 0.06395503841340541,
      "num_tokens": 328009.0,
      "step": 190
    },
    {
      "entropy": 8.346093559265137,
      "epoch": 0.015172145496985022,
      "grad_norm": 0.875,
      "learning_rate": 9.7e-05,
      "loss": 7.642,
      "mean_token_accuracy": 0.0676554836332798,
      "num_tokens": 337414.0,
      "step": 195
    },
    {
      "entropy": 8.347679996490479,
      "epoch": 0.015561174868702588,
      "grad_norm": 0.9140625,
      "learning_rate": 9.95e-05,
      "loss": 7.6106,
      "mean_token_accuracy": 0.0668629266321659,
      "num_tokens": 345419.0,
      "step": 200
    },
    {
      "entropy": 8.210858917236328,
      "epoch": 0.01595020424042015,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000102,
      "loss": 7.5878,
      "mean_token_accuracy": 0.07259676679968834,
      "num_tokens": 353435.0,
      "step": 205
    },
    {
      "entropy": 8.228489208221436,
      "epoch": 0.016339233612137717,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.5984,
      "mean_token_accuracy": 0.07119170576334,
      "num_tokens": 361840.0,
      "step": 210
    },
    {
      "entropy": 8.220147895812989,
      "epoch": 0.016728262983855283,
      "grad_norm": 1.03125,
      "learning_rate": 0.000107,
      "loss": 7.67,
      "mean_token_accuracy": 0.0687949825078249,
      "num_tokens": 370873.0,
      "step": 215
    },
    {
      "entropy": 8.175065612792968,
      "epoch": 0.017117292355572845,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001095,
      "loss": 7.5955,
      "mean_token_accuracy": 0.06820659413933754,
      "num_tokens": 379558.0,
      "step": 220
    },
    {
      "entropy": 8.128092670440674,
      "epoch": 0.01750632172729041,
      "grad_norm": 0.984375,
      "learning_rate": 0.000112,
      "loss": 7.5724,
      "mean_token_accuracy": 0.0715484719723463,
      "num_tokens": 387870.0,
      "step": 225
    },
    {
      "entropy": 8.142854118347168,
      "epoch": 0.017895351099007976,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001145,
      "loss": 7.6264,
      "mean_token_accuracy": 0.06858488656580448,
      "num_tokens": 396442.0,
      "step": 230
    },
    {
      "entropy": 8.081027698516845,
      "epoch": 0.01828438047072554,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.5667,
      "mean_token_accuracy": 0.07171515934169292,
      "num_tokens": 404764.0,
      "step": 235
    },
    {
      "entropy": 8.084673309326172,
      "epoch": 0.018673409842443103,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.5095,
      "mean_token_accuracy": 0.0741395641118288,
      "num_tokens": 413355.0,
      "step": 240
    },
    {
      "entropy": 8.088351917266845,
      "epoch": 0.01906243921416067,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000122,
      "loss": 7.4163,
      "mean_token_accuracy": 0.07463119179010391,
      "num_tokens": 421868.0,
      "step": 245
    },
    {
      "entropy": 8.05438003540039,
      "epoch": 0.019451468585878234,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0001245,
      "loss": 7.4493,
      "mean_token_accuracy": 0.07356020472943783,
      "num_tokens": 431106.0,
      "step": 250
    },
    {
      "entropy": 8.085360860824585,
      "epoch": 0.0198404979575958,
      "grad_norm": 1.03125,
      "learning_rate": 0.000127,
      "loss": 7.5595,
      "mean_token_accuracy": 0.06749935299158097,
      "num_tokens": 439459.0,
      "step": 255
    },
    {
      "entropy": 7.997472143173217,
      "epoch": 0.020229527329313362,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001295,
      "loss": 7.4074,
      "mean_token_accuracy": 0.07336174622178078,
      "num_tokens": 447943.0,
      "step": 260
    },
    {
      "entropy": 7.982085561752319,
      "epoch": 0.020618556701030927,
      "grad_norm": 1.046875,
      "learning_rate": 0.000132,
      "loss": 7.5105,
      "mean_token_accuracy": 0.0743416965007782,
      "num_tokens": 456691.0,
      "step": 265
    },
    {
      "entropy": 7.918378829956055,
      "epoch": 0.021007586072748493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.4348,
      "mean_token_accuracy": 0.074999488145113,
      "num_tokens": 466388.0,
      "step": 270
    },
    {
      "entropy": 8.023542356491088,
      "epoch": 0.02139661544446606,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.4804,
      "mean_token_accuracy": 0.07692455723881722,
      "num_tokens": 474544.0,
      "step": 275
    },
    {
      "entropy": 8.005887222290038,
      "epoch": 0.02178564481618362,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0001395,
      "loss": 7.504,
      "mean_token_accuracy": 0.07702539637684822,
      "num_tokens": 483561.0,
      "step": 280
    },
    {
      "entropy": 7.895681428909302,
      "epoch": 0.022174674187901186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.3428,
      "mean_token_accuracy": 0.07652496919035912,
      "num_tokens": 492036.0,
      "step": 285
    },
    {
      "entropy": 7.991393566131592,
      "epoch": 0.02256370355961875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0001445,
      "loss": 7.457,
      "mean_token_accuracy": 0.07459444478154183,
      "num_tokens": 500978.0,
      "step": 290
    },
    {
      "entropy": 7.994901657104492,
      "epoch": 0.022952732931336317,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000147,
      "loss": 7.4575,
      "mean_token_accuracy": 0.0793934091925621,
      "num_tokens": 509719.0,
      "step": 295
    },
    {
      "entropy": 7.917689561843872,
      "epoch": 0.02334176230305388,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0001495,
      "loss": 7.3986,
      "mean_token_accuracy": 0.07768453806638717,
      "num_tokens": 518422.0,
      "step": 300
    },
    {
      "entropy": 7.879164934158325,
      "epoch": 0.023730791674771445,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000152,
      "loss": 7.4261,
      "mean_token_accuracy": 0.07615871094167233,
      "num_tokens": 527172.0,
      "step": 305
    },
    {
      "entropy": 7.969165182113647,
      "epoch": 0.02411982104648901,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.335,
      "mean_token_accuracy": 0.08110791519284248,
      "num_tokens": 536070.0,
      "step": 310
    },
    {
      "entropy": 7.876494932174682,
      "epoch": 0.024508850418206576,
      "grad_norm": 1.078125,
      "learning_rate": 0.000157,
      "loss": 7.3346,
      "mean_token_accuracy": 0.08139037936925889,
      "num_tokens": 544556.0,
      "step": 315
    },
    {
      "entropy": 7.930807495117188,
      "epoch": 0.024897879789924138,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0001595,
      "loss": 7.3277,
      "mean_token_accuracy": 0.0813445694744587,
      "num_tokens": 553122.0,
      "step": 320
    },
    {
      "entropy": 7.847272491455078,
      "epoch": 0.025286909161641703,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000162,
      "loss": 7.3702,
      "mean_token_accuracy": 0.07732471227645873,
      "num_tokens": 562265.0,
      "step": 325
    },
    {
      "entropy": 7.812228536605835,
      "epoch": 0.02567593853335927,
      "grad_norm": 0.921875,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.3233,
      "mean_token_accuracy": 0.0869563028216362,
      "num_tokens": 570938.0,
      "step": 330
    },
    {
      "entropy": 7.8487406253814695,
      "epoch": 0.026064967905076834,
      "grad_norm": 1.015625,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.2682,
      "mean_token_accuracy": 0.08920462355017662,
      "num_tokens": 578568.0,
      "step": 335
    },
    {
      "entropy": 7.874819087982178,
      "epoch": 0.026453997276794396,
      "grad_norm": 1.109375,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.3216,
      "mean_token_accuracy": 0.0859262391924858,
      "num_tokens": 587013.0,
      "step": 340
    },
    {
      "entropy": 7.878486108779907,
      "epoch": 0.026843026648511962,
      "grad_norm": 1.140625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.3504,
      "mean_token_accuracy": 0.08751785308122635,
      "num_tokens": 595354.0,
      "step": 345
    },
    {
      "entropy": 7.822435426712036,
      "epoch": 0.027232056020229527,
      "grad_norm": 1.078125,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.3303,
      "mean_token_accuracy": 0.08466543480753899,
      "num_tokens": 603971.0,
      "step": 350
    },
    {
      "entropy": 7.758132362365723,
      "epoch": 0.027621085391947093,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000177,
      "loss": 7.2492,
      "mean_token_accuracy": 0.09002660363912582,
      "num_tokens": 612097.0,
      "step": 355
    },
    {
      "entropy": 7.7984230518341064,
      "epoch": 0.028010114763664655,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001795,
      "loss": 7.2982,
      "mean_token_accuracy": 0.08613273575901985,
      "num_tokens": 621445.0,
      "step": 360
    },
    {
      "entropy": 7.769416999816895,
      "epoch": 0.02839914413538222,
      "grad_norm": 1.046875,
      "learning_rate": 0.000182,
      "loss": 7.3609,
      "mean_token_accuracy": 0.08488503694534302,
      "num_tokens": 630041.0,
      "step": 365
    },
    {
      "entropy": 7.751703786849975,
      "epoch": 0.028788173507099786,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001845,
      "loss": 7.1729,
      "mean_token_accuracy": 0.08984732106328011,
      "num_tokens": 638317.0,
      "step": 370
    },
    {
      "entropy": 7.772419214248657,
      "epoch": 0.02917720287881735,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000187,
      "loss": 7.3048,
      "mean_token_accuracy": 0.08521358594298363,
      "num_tokens": 647297.0,
      "step": 375
    },
    {
      "entropy": 7.845639944076538,
      "epoch": 0.029566232250534917,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001895,
      "loss": 7.3325,
      "mean_token_accuracy": 0.08574758470058441,
      "num_tokens": 655309.0,
      "step": 380
    },
    {
      "entropy": 7.704274225234985,
      "epoch": 0.02995526162225248,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000192,
      "loss": 7.2032,
      "mean_token_accuracy": 0.08063894510269165,
      "num_tokens": 663879.0,
      "step": 385
    },
    {
      "entropy": 7.75761194229126,
      "epoch": 0.030344290993970045,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001945,
      "loss": 7.1131,
      "mean_token_accuracy": 0.09121661111712456,
      "num_tokens": 672913.0,
      "step": 390
    },
    {
      "entropy": 7.677809333801269,
      "epoch": 0.03073332036568761,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.2217,
      "mean_token_accuracy": 0.08521161302924156,
      "num_tokens": 682282.0,
      "step": 395
    },
    {
      "entropy": 7.67137598991394,
      "epoch": 0.031122349737405176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.2312,
      "mean_token_accuracy": 0.08108669146895409,
      "num_tokens": 691432.0,
      "step": 400
    },
    {
      "entropy": 7.725798177719116,
      "epoch": 0.03151137910912274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000202,
      "loss": 7.2266,
      "mean_token_accuracy": 0.08599605634808541,
      "num_tokens": 700639.0,
      "step": 405
    },
    {
      "entropy": 7.75080361366272,
      "epoch": 0.0319004084808403,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.2862,
      "mean_token_accuracy": 0.08474635109305381,
      "num_tokens": 709200.0,
      "step": 410
    },
    {
      "entropy": 7.598263549804687,
      "epoch": 0.03228943785255787,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000207,
      "loss": 7.1093,
      "mean_token_accuracy": 0.09120492935180664,
      "num_tokens": 717266.0,
      "step": 415
    },
    {
      "entropy": 7.665576219558716,
      "epoch": 0.032678467224275434,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002095,
      "loss": 7.1208,
      "mean_token_accuracy": 0.0892751768231392,
      "num_tokens": 725799.0,
      "step": 420
    },
    {
      "entropy": 7.653212070465088,
      "epoch": 0.033067496595993,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000212,
      "loss": 7.2111,
      "mean_token_accuracy": 0.08835389018058777,
      "num_tokens": 734375.0,
      "step": 425
    },
    {
      "entropy": 7.74424262046814,
      "epoch": 0.033456525967710565,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002145,
      "loss": 7.2555,
      "mean_token_accuracy": 0.08766974285244941,
      "num_tokens": 743453.0,
      "step": 430
    },
    {
      "entropy": 7.593816137313842,
      "epoch": 0.033845555339428124,
      "grad_norm": 0.9375,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.1001,
      "mean_token_accuracy": 0.09561214223504066,
      "num_tokens": 752443.0,
      "step": 435
    },
    {
      "entropy": 7.626216220855713,
      "epoch": 0.03423458471114569,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002195,
      "loss": 7.0988,
      "mean_token_accuracy": 0.08553416356444359,
      "num_tokens": 761294.0,
      "step": 440
    },
    {
      "entropy": 7.621181488037109,
      "epoch": 0.034623614082863255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000222,
      "loss": 7.0642,
      "mean_token_accuracy": 0.09563054814934731,
      "num_tokens": 769489.0,
      "step": 445
    },
    {
      "entropy": 7.661951541900635,
      "epoch": 0.03501264345458082,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002245,
      "loss": 7.1244,
      "mean_token_accuracy": 0.09401247352361679,
      "num_tokens": 777836.0,
      "step": 450
    },
    {
      "entropy": 7.584022903442383,
      "epoch": 0.035401672826298386,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.1516,
      "mean_token_accuracy": 0.08489634171128273,
      "num_tokens": 786543.0,
      "step": 455
    },
    {
      "entropy": 7.590439224243164,
      "epoch": 0.03579070219801595,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.1911,
      "mean_token_accuracy": 0.09159761071205139,
      "num_tokens": 794525.0,
      "step": 460
    },
    {
      "entropy": 7.574541854858398,
      "epoch": 0.03617973156973352,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.0976,
      "mean_token_accuracy": 0.0944310788065195,
      "num_tokens": 803537.0,
      "step": 465
    },
    {
      "entropy": 7.611633348464966,
      "epoch": 0.03656876094145108,
      "grad_norm": 1.0,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.0519,
      "mean_token_accuracy": 0.09074848592281341,
      "num_tokens": 811706.0,
      "step": 470
    },
    {
      "entropy": 7.562524032592774,
      "epoch": 0.03695779031316864,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000237,
      "loss": 7.1067,
      "mean_token_accuracy": 0.09484568610787392,
      "num_tokens": 820182.0,
      "step": 475
    },
    {
      "entropy": 7.572012710571289,
      "epoch": 0.03734681968488621,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002395,
      "loss": 7.0488,
      "mean_token_accuracy": 0.09293625578284263,
      "num_tokens": 829808.0,
      "step": 480
    },
    {
      "entropy": 7.545894145965576,
      "epoch": 0.03773584905660377,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000242,
      "loss": 6.9895,
      "mean_token_accuracy": 0.09443094059824944,
      "num_tokens": 838838.0,
      "step": 485
    },
    {
      "entropy": 7.557442235946655,
      "epoch": 0.03812487842832134,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002445,
      "loss": 7.0937,
      "mean_token_accuracy": 0.09211538061499595,
      "num_tokens": 847800.0,
      "step": 490
    },
    {
      "entropy": 7.60601692199707,
      "epoch": 0.0385139078000389,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000247,
      "loss": 7.1669,
      "mean_token_accuracy": 0.09093088284134865,
      "num_tokens": 856503.0,
      "step": 495
    },
    {
      "entropy": 7.475769996643066,
      "epoch": 0.03890293717175647,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002495,
      "loss": 7.05,
      "mean_token_accuracy": 0.09106085672974587,
      "num_tokens": 865245.0,
      "step": 500
    },
    {
      "entropy": 7.588863754272461,
      "epoch": 0.039291966543474034,
      "grad_norm": 1.1875,
      "learning_rate": 0.000252,
      "loss": 7.0721,
      "mean_token_accuracy": 0.09027522280812264,
      "num_tokens": 874485.0,
      "step": 505
    },
    {
      "entropy": 7.418361043930053,
      "epoch": 0.0396809959151916,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002545,
      "loss": 6.9375,
      "mean_token_accuracy": 0.09956656396389008,
      "num_tokens": 883301.0,
      "step": 510
    },
    {
      "entropy": 7.408757543563842,
      "epoch": 0.04007002528690916,
      "grad_norm": 1.125,
      "learning_rate": 0.000257,
      "loss": 6.9846,
      "mean_token_accuracy": 0.0950555182993412,
      "num_tokens": 892387.0,
      "step": 515
    },
    {
      "entropy": 7.48194580078125,
      "epoch": 0.040459054658626724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002595,
      "loss": 6.9709,
      "mean_token_accuracy": 0.09912419691681862,
      "num_tokens": 901522.0,
      "step": 520
    },
    {
      "entropy": 7.441246747970581,
      "epoch": 0.04084808403034429,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000262,
      "loss": 6.9292,
      "mean_token_accuracy": 0.0985661618411541,
      "num_tokens": 909848.0,
      "step": 525
    },
    {
      "entropy": 7.562440967559814,
      "epoch": 0.041237113402061855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.164,
      "mean_token_accuracy": 0.08993402048945427,
      "num_tokens": 919079.0,
      "step": 530
    },
    {
      "entropy": 7.420489549636841,
      "epoch": 0.04162614277377942,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.9063,
      "mean_token_accuracy": 0.09894024208188057,
      "num_tokens": 927807.0,
      "step": 535
    },
    {
      "entropy": 7.408330631256104,
      "epoch": 0.042015172145496986,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.0018,
      "mean_token_accuracy": 0.09915620237588882,
      "num_tokens": 936498.0,
      "step": 540
    },
    {
      "entropy": 7.487834930419922,
      "epoch": 0.04240420151721455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.0738,
      "mean_token_accuracy": 0.0998169869184494,
      "num_tokens": 945436.0,
      "step": 545
    },
    {
      "entropy": 7.528131818771362,
      "epoch": 0.04279323088893212,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002745,
      "loss": 6.9875,
      "mean_token_accuracy": 0.09678835868835449,
      "num_tokens": 953990.0,
      "step": 550
    },
    {
      "entropy": 7.37784481048584,
      "epoch": 0.04318226026064968,
      "grad_norm": 1.234375,
      "learning_rate": 0.000277,
      "loss": 6.9245,
      "mean_token_accuracy": 0.09856464117765426,
      "num_tokens": 962008.0,
      "step": 555
    },
    {
      "entropy": 7.438614797592163,
      "epoch": 0.04357128963236724,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002795,
      "loss": 7.0788,
      "mean_token_accuracy": 0.09496510401368141,
      "num_tokens": 970935.0,
      "step": 560
    },
    {
      "entropy": 7.521619749069214,
      "epoch": 0.04396031900408481,
      "grad_norm": 1.125,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.0371,
      "mean_token_accuracy": 0.08908204212784768,
      "num_tokens": 979492.0,
      "step": 565
    },
    {
      "entropy": 7.361342287063598,
      "epoch": 0.04434934837580237,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002845,
      "loss": 6.9321,
      "mean_token_accuracy": 0.09547191262245178,
      "num_tokens": 987866.0,
      "step": 570
    },
    {
      "entropy": 7.397969961166382,
      "epoch": 0.04473837774751994,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000287,
      "loss": 6.9804,
      "mean_token_accuracy": 0.09910744205117225,
      "num_tokens": 996760.0,
      "step": 575
    },
    {
      "entropy": 7.447316217422485,
      "epoch": 0.0451274071192375,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002895,
      "loss": 6.9028,
      "mean_token_accuracy": 0.10264215171337128,
      "num_tokens": 1005432.0,
      "step": 580
    },
    {
      "entropy": 7.396947145462036,
      "epoch": 0.04551643649095507,
      "grad_norm": 1.171875,
      "learning_rate": 0.000292,
      "loss": 6.9787,
      "mean_token_accuracy": 0.09953495115041733,
      "num_tokens": 1013729.0,
      "step": 585
    },
    {
      "entropy": 7.278457880020142,
      "epoch": 0.045905465862672634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002945,
      "loss": 6.8221,
      "mean_token_accuracy": 0.10534313321113586,
      "num_tokens": 1022485.0,
      "step": 590
    },
    {
      "entropy": 7.3537023067474365,
      "epoch": 0.0462944952343902,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000297,
      "loss": 6.8869,
      "mean_token_accuracy": 0.0984025850892067,
      "num_tokens": 1030523.0,
      "step": 595
    },
    {
      "entropy": 7.419099044799805,
      "epoch": 0.04668352460610776,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002995,
      "loss": 6.99,
      "mean_token_accuracy": 0.09814063236117362,
      "num_tokens": 1039268.0,
      "step": 600
    },
    {
      "entropy": 7.301190805435181,
      "epoch": 0.047072553977825324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000302,
      "loss": 6.8802,
      "mean_token_accuracy": 0.09921084493398666,
      "num_tokens": 1048590.0,
      "step": 605
    },
    {
      "entropy": 7.268998908996582,
      "epoch": 0.04746158334954289,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003045,
      "loss": 6.811,
      "mean_token_accuracy": 0.10243154391646385,
      "num_tokens": 1057232.0,
      "step": 610
    },
    {
      "entropy": 7.301156949996948,
      "epoch": 0.047850612721260455,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000307,
      "loss": 6.8863,
      "mean_token_accuracy": 0.09728901833295822,
      "num_tokens": 1066404.0,
      "step": 615
    },
    {
      "entropy": 7.226704359054565,
      "epoch": 0.04823964209297802,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003095,
      "loss": 6.957,
      "mean_token_accuracy": 0.09339571669697762,
      "num_tokens": 1075093.0,
      "step": 620
    },
    {
      "entropy": 7.38017783164978,
      "epoch": 0.048628671464695586,
      "grad_norm": 1.0625,
      "learning_rate": 0.000312,
      "loss": 6.864,
      "mean_token_accuracy": 0.09610054194927216,
      "num_tokens": 1084454.0,
      "step": 625
    },
    {
      "entropy": 7.280154037475586,
      "epoch": 0.04901770083641315,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003145,
      "loss": 6.7866,
      "mean_token_accuracy": 0.10086495503783226,
      "num_tokens": 1093452.0,
      "step": 630
    },
    {
      "entropy": 7.33285608291626,
      "epoch": 0.04940673020813072,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000317,
      "loss": 6.807,
      "mean_token_accuracy": 0.10476917698979378,
      "num_tokens": 1101814.0,
      "step": 635
    },
    {
      "entropy": 7.184072637557984,
      "epoch": 0.049795759579848276,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003195,
      "loss": 6.8889,
      "mean_token_accuracy": 0.10318943634629249,
      "num_tokens": 1110386.0,
      "step": 640
    },
    {
      "entropy": 7.352627325057983,
      "epoch": 0.05018478895156584,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000322,
      "loss": 6.9831,
      "mean_token_accuracy": 0.0920788213610649,
      "num_tokens": 1120141.0,
      "step": 645
    },
    {
      "entropy": 7.324120092391968,
      "epoch": 0.05057381832328341,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.9227,
      "mean_token_accuracy": 0.09927786216139793,
      "num_tokens": 1129920.0,
      "step": 650
    },
    {
      "entropy": 7.171842193603515,
      "epoch": 0.05096284769500097,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.7957,
      "mean_token_accuracy": 0.10289519429206848,
      "num_tokens": 1138518.0,
      "step": 655
    },
    {
      "entropy": 7.32971568107605,
      "epoch": 0.05135187706671854,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.8899,
      "mean_token_accuracy": 0.1004197433590889,
      "num_tokens": 1147113.0,
      "step": 660
    },
    {
      "entropy": 7.201927232742309,
      "epoch": 0.0517409064384361,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.7435,
      "mean_token_accuracy": 0.10898910462856293,
      "num_tokens": 1155886.0,
      "step": 665
    },
    {
      "entropy": 7.199523496627807,
      "epoch": 0.05212993581015367,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.8999,
      "mean_token_accuracy": 0.09887881278991699,
      "num_tokens": 1164200.0,
      "step": 670
    },
    {
      "entropy": 7.209548091888427,
      "epoch": 0.052518965181871234,
      "grad_norm": 1.0625,
      "learning_rate": 0.000337,
      "loss": 6.7543,
      "mean_token_accuracy": 0.10513453632593155,
      "num_tokens": 1172720.0,
      "step": 675
    },
    {
      "entropy": 7.177516746520996,
      "epoch": 0.05290799455358879,
      "grad_norm": 1.125,
      "learning_rate": 0.0003395,
      "loss": 6.7115,
      "mean_token_accuracy": 0.09829143732786179,
      "num_tokens": 1181225.0,
      "step": 680
    },
    {
      "entropy": 7.2649918556213375,
      "epoch": 0.05329702392530636,
      "grad_norm": 1.09375,
      "learning_rate": 0.000342,
      "loss": 6.9086,
      "mean_token_accuracy": 0.09883515611290931,
      "num_tokens": 1190070.0,
      "step": 685
    },
    {
      "entropy": 7.220799922943115,
      "epoch": 0.053686053297023924,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.7985,
      "mean_token_accuracy": 0.1037118412554264,
      "num_tokens": 1198995.0,
      "step": 690
    },
    {
      "entropy": 7.23320894241333,
      "epoch": 0.05407508266874149,
      "grad_norm": 1.140625,
      "learning_rate": 0.000347,
      "loss": 6.8442,
      "mean_token_accuracy": 0.1092130720615387,
      "num_tokens": 1207664.0,
      "step": 695
    },
    {
      "entropy": 7.272186803817749,
      "epoch": 0.054464112040459055,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003495,
      "loss": 6.7473,
      "mean_token_accuracy": 0.11652912721037864,
      "num_tokens": 1216232.0,
      "step": 700
    },
    {
      "entropy": 7.137939691543579,
      "epoch": 0.05485314141217662,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000352,
      "loss": 6.8328,
      "mean_token_accuracy": 0.1046180248260498,
      "num_tokens": 1224106.0,
      "step": 705
    },
    {
      "entropy": 7.112717533111573,
      "epoch": 0.055242170783894186,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003545,
      "loss": 6.7241,
      "mean_token_accuracy": 0.1101645365357399,
      "num_tokens": 1232514.0,
      "step": 710
    },
    {
      "entropy": 7.2445765972137455,
      "epoch": 0.05563120015561175,
      "grad_norm": 1.234375,
      "learning_rate": 0.000357,
      "loss": 6.7763,
      "mean_token_accuracy": 0.10698423609137535,
      "num_tokens": 1240178.0,
      "step": 715
    },
    {
      "entropy": 7.076436710357666,
      "epoch": 0.05602022952732931,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003595,
      "loss": 6.6795,
      "mean_token_accuracy": 0.10781070590019226,
      "num_tokens": 1248635.0,
      "step": 720
    },
    {
      "entropy": 7.209420871734619,
      "epoch": 0.056409258899046875,
      "grad_norm": 1.28125,
      "learning_rate": 0.000362,
      "loss": 6.7874,
      "mean_token_accuracy": 0.1086701013147831,
      "num_tokens": 1256389.0,
      "step": 725
    },
    {
      "entropy": 7.077723169326783,
      "epoch": 0.05679828827076444,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003645,
      "loss": 6.7275,
      "mean_token_accuracy": 0.10797203481197357,
      "num_tokens": 1264035.0,
      "step": 730
    },
    {
      "entropy": 7.17639570236206,
      "epoch": 0.057187317642482006,
      "grad_norm": 1.203125,
      "learning_rate": 0.000367,
      "loss": 6.7417,
      "mean_token_accuracy": 0.10852355882525444,
      "num_tokens": 1272941.0,
      "step": 735
    },
    {
      "entropy": 6.992896270751953,
      "epoch": 0.05757634701419957,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003695,
      "loss": 6.719,
      "mean_token_accuracy": 0.10654684156179428,
      "num_tokens": 1281767.0,
      "step": 740
    },
    {
      "entropy": 7.257743215560913,
      "epoch": 0.05796537638591714,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000372,
      "loss": 6.8008,
      "mean_token_accuracy": 0.10871746987104416,
      "num_tokens": 1290049.0,
      "step": 745
    },
    {
      "entropy": 7.025216960906983,
      "epoch": 0.0583544057576347,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003745,
      "loss": 6.6344,
      "mean_token_accuracy": 0.11401695236563683,
      "num_tokens": 1299032.0,
      "step": 750
    },
    {
      "entropy": 7.10672435760498,
      "epoch": 0.05874343512935227,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000377,
      "loss": 6.8797,
      "mean_token_accuracy": 0.10258082747459411,
      "num_tokens": 1308508.0,
      "step": 755
    },
    {
      "entropy": 7.206971597671509,
      "epoch": 0.059132464501069834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003795,
      "loss": 6.7266,
      "mean_token_accuracy": 0.10653159618377686,
      "num_tokens": 1317044.0,
      "step": 760
    },
    {
      "entropy": 7.129873371124267,
      "epoch": 0.05952149387278739,
      "grad_norm": 1.09375,
      "learning_rate": 0.000382,
      "loss": 6.7401,
      "mean_token_accuracy": 0.10757752284407615,
      "num_tokens": 1325554.0,
      "step": 765
    },
    {
      "entropy": 7.057963895797729,
      "epoch": 0.05991052324450496,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003845,
      "loss": 6.6354,
      "mean_token_accuracy": 0.10638404935598374,
      "num_tokens": 1334201.0,
      "step": 770
    },
    {
      "entropy": 7.028696727752686,
      "epoch": 0.060299552616222524,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.6702,
      "mean_token_accuracy": 0.11282677426934243,
      "num_tokens": 1343061.0,
      "step": 775
    },
    {
      "entropy": 7.1010072231292725,
      "epoch": 0.06068858198794009,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6186,
      "mean_token_accuracy": 0.11468041315674782,
      "num_tokens": 1352262.0,
      "step": 780
    },
    {
      "entropy": 7.018054246902466,
      "epoch": 0.061077611359657655,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.6796,
      "mean_token_accuracy": 0.10959815979003906,
      "num_tokens": 1361400.0,
      "step": 785
    },
    {
      "entropy": 7.052734136581421,
      "epoch": 0.06146664073137522,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.7896,
      "mean_token_accuracy": 0.10606501176953316,
      "num_tokens": 1369850.0,
      "step": 790
    },
    {
      "entropy": 7.069890785217285,
      "epoch": 0.061855670103092786,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.6306,
      "mean_token_accuracy": 0.11038519069552422,
      "num_tokens": 1378342.0,
      "step": 795
    },
    {
      "entropy": 7.049418497085571,
      "epoch": 0.06224469947481035,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003995,
      "loss": 6.7302,
      "mean_token_accuracy": 0.1076231300830841,
      "num_tokens": 1386981.0,
      "step": 800
    },
    {
      "entropy": 7.077835321426392,
      "epoch": 0.06263372884652792,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000402,
      "loss": 6.8099,
      "mean_token_accuracy": 0.10120006203651429,
      "num_tokens": 1395801.0,
      "step": 805
    },
    {
      "entropy": 7.206391859054565,
      "epoch": 0.06302275821824548,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004045,
      "loss": 6.7163,
      "mean_token_accuracy": 0.1102595366537571,
      "num_tokens": 1403715.0,
      "step": 810
    },
    {
      "entropy": 7.088693380355835,
      "epoch": 0.06341178758996305,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.7922,
      "mean_token_accuracy": 0.10654390528798104,
      "num_tokens": 1413199.0,
      "step": 815
    },
    {
      "entropy": 7.05805082321167,
      "epoch": 0.0638008169616806,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004095,
      "loss": 6.7592,
      "mean_token_accuracy": 0.10669645592570305,
      "num_tokens": 1421918.0,
      "step": 820
    },
    {
      "entropy": 7.013793468475342,
      "epoch": 0.06418984633339817,
      "grad_norm": 1.09375,
      "learning_rate": 0.000412,
      "loss": 6.6581,
      "mean_token_accuracy": 0.1045446127653122,
      "num_tokens": 1430625.0,
      "step": 825
    },
    {
      "entropy": 6.975633525848389,
      "epoch": 0.06457887570511574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004145,
      "loss": 6.6945,
      "mean_token_accuracy": 0.1070571556687355,
      "num_tokens": 1439158.0,
      "step": 830
    },
    {
      "entropy": 6.98923978805542,
      "epoch": 0.0649679050768333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000417,
      "loss": 6.6591,
      "mean_token_accuracy": 0.11569598615169525,
      "num_tokens": 1448266.0,
      "step": 835
    },
    {
      "entropy": 7.057836484909058,
      "epoch": 0.06535693444855087,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004195,
      "loss": 6.7327,
      "mean_token_accuracy": 0.11181987822055817,
      "num_tokens": 1457322.0,
      "step": 840
    },
    {
      "entropy": 7.135942649841309,
      "epoch": 0.06574596382026843,
      "grad_norm": 1.03125,
      "learning_rate": 0.000422,
      "loss": 6.8529,
      "mean_token_accuracy": 0.10292978584766388,
      "num_tokens": 1466120.0,
      "step": 845
    },
    {
      "entropy": 7.01804838180542,
      "epoch": 0.066134993191986,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004245,
      "loss": 6.6053,
      "mean_token_accuracy": 0.11059460267424584,
      "num_tokens": 1474965.0,
      "step": 850
    },
    {
      "entropy": 6.987716579437256,
      "epoch": 0.06652402256370356,
      "grad_norm": 1.15625,
      "learning_rate": 0.000427,
      "loss": 6.6871,
      "mean_token_accuracy": 0.10879706963896751,
      "num_tokens": 1483483.0,
      "step": 855
    },
    {
      "entropy": 7.070033025741577,
      "epoch": 0.06691305193542113,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004295,
      "loss": 6.6423,
      "mean_token_accuracy": 0.10447335988283157,
      "num_tokens": 1492313.0,
      "step": 860
    },
    {
      "entropy": 6.95804386138916,
      "epoch": 0.06730208130713869,
      "grad_norm": 1.078125,
      "learning_rate": 0.000432,
      "loss": 6.5724,
      "mean_token_accuracy": 0.11189851835370064,
      "num_tokens": 1501081.0,
      "step": 865
    },
    {
      "entropy": 6.913361120223999,
      "epoch": 0.06769111067885625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004345,
      "loss": 6.5954,
      "mean_token_accuracy": 0.10529705733060837,
      "num_tokens": 1510169.0,
      "step": 870
    },
    {
      "entropy": 6.990754508972168,
      "epoch": 0.06808014005057382,
      "grad_norm": 1.171875,
      "learning_rate": 0.000437,
      "loss": 6.5564,
      "mean_token_accuracy": 0.10811040028929711,
      "num_tokens": 1519546.0,
      "step": 875
    },
    {
      "entropy": 6.92680926322937,
      "epoch": 0.06846916942229138,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004395,
      "loss": 6.6583,
      "mean_token_accuracy": 0.10640382617712021,
      "num_tokens": 1529104.0,
      "step": 880
    },
    {
      "entropy": 6.795887804031372,
      "epoch": 0.06885819879400895,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000442,
      "loss": 6.6051,
      "mean_token_accuracy": 0.1192264623939991,
      "num_tokens": 1537033.0,
      "step": 885
    },
    {
      "entropy": 7.02658314704895,
      "epoch": 0.06924722816572651,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004445,
      "loss": 6.6247,
      "mean_token_accuracy": 0.11063896790146828,
      "num_tokens": 1545059.0,
      "step": 890
    },
    {
      "entropy": 7.039566946029663,
      "epoch": 0.06963625753744408,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000447,
      "loss": 6.6317,
      "mean_token_accuracy": 0.10993520393967629,
      "num_tokens": 1553784.0,
      "step": 895
    },
    {
      "entropy": 6.85557713508606,
      "epoch": 0.07002528690916164,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.5602,
      "mean_token_accuracy": 0.11566279828548431,
      "num_tokens": 1562059.0,
      "step": 900
    },
    {
      "entropy": 6.923605155944824,
      "epoch": 0.0704143162808792,
      "grad_norm": 1.125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.6166,
      "mean_token_accuracy": 0.10865532159805298,
      "num_tokens": 1570061.0,
      "step": 905
    },
    {
      "entropy": 6.915492105484009,
      "epoch": 0.07080334565259677,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.5864,
      "mean_token_accuracy": 0.11283667013049126,
      "num_tokens": 1578261.0,
      "step": 910
    },
    {
      "entropy": 6.870081186294556,
      "epoch": 0.07119237502431433,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.5254,
      "mean_token_accuracy": 0.11373848989605903,
      "num_tokens": 1586260.0,
      "step": 915
    },
    {
      "entropy": 6.957163333892822,
      "epoch": 0.0715814043960319,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.7037,
      "mean_token_accuracy": 0.10588114485144615,
      "num_tokens": 1595234.0,
      "step": 920
    },
    {
      "entropy": 6.966896820068359,
      "epoch": 0.07197043376774946,
      "grad_norm": 1.25,
      "learning_rate": 0.000462,
      "loss": 6.6038,
      "mean_token_accuracy": 0.10876048058271408,
      "num_tokens": 1604212.0,
      "step": 925
    },
    {
      "entropy": 6.890627479553222,
      "epoch": 0.07235946313946703,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004645,
      "loss": 6.6407,
      "mean_token_accuracy": 0.11045560985803604,
      "num_tokens": 1612925.0,
      "step": 930
    },
    {
      "entropy": 6.871489381790161,
      "epoch": 0.07274849251118459,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000467,
      "loss": 6.5285,
      "mean_token_accuracy": 0.111960668861866,
      "num_tokens": 1621804.0,
      "step": 935
    },
    {
      "entropy": 6.901230573654175,
      "epoch": 0.07313752188290217,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004695,
      "loss": 6.5111,
      "mean_token_accuracy": 0.1133945107460022,
      "num_tokens": 1630943.0,
      "step": 940
    },
    {
      "entropy": 6.869029235839844,
      "epoch": 0.07352655125461972,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000472,
      "loss": 6.4213,
      "mean_token_accuracy": 0.11604218557476997,
      "num_tokens": 1638899.0,
      "step": 945
    },
    {
      "entropy": 6.868594026565551,
      "epoch": 0.07391558062633728,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004745,
      "loss": 6.6351,
      "mean_token_accuracy": 0.11411484777927398,
      "num_tokens": 1647485.0,
      "step": 950
    },
    {
      "entropy": 6.925864124298096,
      "epoch": 0.07430460999805485,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000477,
      "loss": 6.6023,
      "mean_token_accuracy": 0.11547646671533585,
      "num_tokens": 1656290.0,
      "step": 955
    },
    {
      "entropy": 6.93121566772461,
      "epoch": 0.07469363936977241,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004795,
      "loss": 6.611,
      "mean_token_accuracy": 0.1147051364183426,
      "num_tokens": 1665474.0,
      "step": 960
    },
    {
      "entropy": 6.733439493179321,
      "epoch": 0.07508266874148999,
      "grad_norm": 1.25,
      "learning_rate": 0.000482,
      "loss": 6.4883,
      "mean_token_accuracy": 0.11983574107289315,
      "num_tokens": 1673599.0,
      "step": 965
    },
    {
      "entropy": 6.897821712493896,
      "epoch": 0.07547169811320754,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004845,
      "loss": 6.4739,
      "mean_token_accuracy": 0.11102246791124344,
      "num_tokens": 1683282.0,
      "step": 970
    },
    {
      "entropy": 6.931991624832153,
      "epoch": 0.07586072748492512,
      "grad_norm": 1.171875,
      "learning_rate": 0.000487,
      "loss": 6.5991,
      "mean_token_accuracy": 0.11667483672499657,
      "num_tokens": 1691094.0,
      "step": 975
    },
    {
      "entropy": 6.752887296676636,
      "epoch": 0.07624975685664268,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004895,
      "loss": 6.4645,
      "mean_token_accuracy": 0.11685017794370652,
      "num_tokens": 1700053.0,
      "step": 980
    },
    {
      "entropy": 6.807966899871826,
      "epoch": 0.07663878622836025,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000492,
      "loss": 6.3961,
      "mean_token_accuracy": 0.11701971888542176,
      "num_tokens": 1707732.0,
      "step": 985
    },
    {
      "entropy": 6.825413990020752,
      "epoch": 0.0770278156000778,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004945,
      "loss": 6.6425,
      "mean_token_accuracy": 0.10670604258775711,
      "num_tokens": 1716406.0,
      "step": 990
    },
    {
      "entropy": 7.002917194366455,
      "epoch": 0.07741684497179536,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497,
      "loss": 6.6635,
      "mean_token_accuracy": 0.10928485170006752,
      "num_tokens": 1725832.0,
      "step": 995
    },
    {
      "entropy": 6.912725305557251,
      "epoch": 0.07780587434351294,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995,
      "loss": 6.576,
      "mean_token_accuracy": 0.11019504964351653,
      "num_tokens": 1734853.0,
      "step": 1000
    },
    {
      "entropy": 6.769201183319092,
      "epoch": 0.0781949037152305,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999999989075146,
      "loss": 6.5842,
      "mean_token_accuracy": 0.11545023322105408,
      "num_tokens": 1744240.0,
      "step": 1005
    },
    {
      "entropy": 6.9675556182861325,
      "epoch": 0.07858393308694807,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999999944692927,
      "loss": 6.6019,
      "mean_token_accuracy": 0.1125156231224537,
      "num_tokens": 1753334.0,
      "step": 1010
    },
    {
      "entropy": 6.70109372138977,
      "epoch": 0.07897296245866563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499999986617054,
      "loss": 6.4932,
      "mean_token_accuracy": 0.11150014474987983,
      "num_tokens": 1762032.0,
      "step": 1015
    },
    {
      "entropy": 6.8338549613952635,
      "epoch": 0.0793619918303832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999753507986,
      "loss": 6.4863,
      "mean_token_accuracy": 0.11209132522344589,
      "num_tokens": 1770508.0,
      "step": 1020
    },
    {
      "entropy": 6.721220541000366,
      "epoch": 0.07975102120210076,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999999606705267,
      "loss": 6.3881,
      "mean_token_accuracy": 0.12179349660873413,
      "num_tokens": 1780167.0,
      "step": 1025
    },
    {
      "entropy": 6.887646627426148,
      "epoch": 0.08014005057381832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999425762385,
      "loss": 6.6013,
      "mean_token_accuracy": 0.10411422997713089,
      "num_tokens": 1788809.0,
      "step": 1030
    },
    {
      "entropy": 6.785115766525268,
      "epoch": 0.08052907994553589,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999999210679344,
      "loss": 6.4975,
      "mean_token_accuracy": 0.1171350821852684,
      "num_tokens": 1797445.0,
      "step": 1035
    },
    {
      "entropy": 6.8339056968688965,
      "epoch": 0.08091810931725345,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999998961456145,
      "loss": 6.5205,
      "mean_token_accuracy": 0.11023695170879363,
      "num_tokens": 1806488.0,
      "step": 1040
    },
    {
      "entropy": 6.893251085281372,
      "epoch": 0.08130713868897102,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999998678092794,
      "loss": 6.5876,
      "mean_token_accuracy": 0.11547244265675545,
      "num_tokens": 1815692.0,
      "step": 1045
    },
    {
      "entropy": 6.8416530132293705,
      "epoch": 0.08169616806068858,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999998360589293,
      "loss": 6.5062,
      "mean_token_accuracy": 0.1161237508058548,
      "num_tokens": 1824599.0,
      "step": 1050
    },
    {
      "entropy": 6.88067593574524,
      "epoch": 0.08208519743240615,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499999800894565,
      "loss": 6.652,
      "mean_token_accuracy": 0.11225388348102569,
      "num_tokens": 1833411.0,
      "step": 1055
    },
    {
      "entropy": 6.79389796257019,
      "epoch": 0.08247422680412371,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999997623161866,
      "loss": 6.4656,
      "mean_token_accuracy": 0.11337645724415779,
      "num_tokens": 1842177.0,
      "step": 1060
    },
    {
      "entropy": 6.698552227020263,
      "epoch": 0.08286325617584128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999997203237951,
      "loss": 6.4423,
      "mean_token_accuracy": 0.12076986059546471,
      "num_tokens": 1850176.0,
      "step": 1065
    },
    {
      "entropy": 6.799434804916382,
      "epoch": 0.08325228554755884,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499999674917391,
      "loss": 6.5019,
      "mean_token_accuracy": 0.11561930030584336,
      "num_tokens": 1859464.0,
      "step": 1070
    },
    {
      "entropy": 6.655920028686523,
      "epoch": 0.0836413149192764,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999996260969748,
      "loss": 6.4004,
      "mean_token_accuracy": 0.11484229564666748,
      "num_tokens": 1867733.0,
      "step": 1075
    },
    {
      "entropy": 6.747368717193604,
      "epoch": 0.08403034429099397,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999995738625475,
      "loss": 6.3873,
      "mean_token_accuracy": 0.12014570608735084,
      "num_tokens": 1876010.0,
      "step": 1080
    },
    {
      "entropy": 6.721405076980591,
      "epoch": 0.08441937366271153,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999995182141099,
      "loss": 6.5431,
      "mean_token_accuracy": 0.11369292438030243,
      "num_tokens": 1884906.0,
      "step": 1085
    },
    {
      "entropy": 6.796774196624756,
      "epoch": 0.0848084030344291,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999994591516626,
      "loss": 6.4584,
      "mean_token_accuracy": 0.11704446524381637,
      "num_tokens": 1893464.0,
      "step": 1090
    },
    {
      "entropy": 6.746003484725952,
      "epoch": 0.08519743240614666,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999993966752066,
      "loss": 6.5049,
      "mean_token_accuracy": 0.11562549471855163,
      "num_tokens": 1902769.0,
      "step": 1095
    },
    {
      "entropy": 6.790870237350464,
      "epoch": 0.08558646177786423,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499999330784743,
      "loss": 6.4273,
      "mean_token_accuracy": 0.11782810240983962,
      "num_tokens": 1911968.0,
      "step": 1100
    },
    {
      "entropy": 6.765281391143799,
      "epoch": 0.08597549114958179,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999992614802725,
      "loss": 6.4144,
      "mean_token_accuracy": 0.11969999745488166,
      "num_tokens": 1920063.0,
      "step": 1105
    },
    {
      "entropy": 6.616898441314698,
      "epoch": 0.08636452052129936,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999991887617966,
      "loss": 6.3855,
      "mean_token_accuracy": 0.11925996840000153,
      "num_tokens": 1929192.0,
      "step": 1110
    },
    {
      "entropy": 6.807584524154663,
      "epoch": 0.08675354989301692,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999991126293158,
      "loss": 6.5924,
      "mean_token_accuracy": 0.11094593480229378,
      "num_tokens": 1939082.0,
      "step": 1115
    },
    {
      "entropy": 6.590944242477417,
      "epoch": 0.08714257926473448,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999990330828318,
      "loss": 6.4607,
      "mean_token_accuracy": 0.11833893656730651,
      "num_tokens": 1948313.0,
      "step": 1120
    },
    {
      "entropy": 6.887479496002197,
      "epoch": 0.08753160863645205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999989501223456,
      "loss": 6.4559,
      "mean_token_accuracy": 0.11990888267755509,
      "num_tokens": 1957035.0,
      "step": 1125
    },
    {
      "entropy": 6.613681793212891,
      "epoch": 0.08792063800816961,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999988637478584,
      "loss": 6.4451,
      "mean_token_accuracy": 0.11189156025648117,
      "num_tokens": 1966041.0,
      "step": 1130
    },
    {
      "entropy": 6.8127364158630375,
      "epoch": 0.08830966737988719,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999987739593716,
      "loss": 6.4775,
      "mean_token_accuracy": 0.10807354152202606,
      "num_tokens": 1975019.0,
      "step": 1135
    },
    {
      "entropy": 6.693046903610229,
      "epoch": 0.08869869675160474,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999986807568865,
      "loss": 6.4086,
      "mean_token_accuracy": 0.12410445287823677,
      "num_tokens": 1982930.0,
      "step": 1140
    },
    {
      "entropy": 6.738013505935669,
      "epoch": 0.08908772612332232,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999985841404045,
      "loss": 6.5192,
      "mean_token_accuracy": 0.11475513577461242,
      "num_tokens": 1991888.0,
      "step": 1145
    },
    {
      "entropy": 6.7153209209442135,
      "epoch": 0.08947675549503988,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999984841099271,
      "loss": 6.3537,
      "mean_token_accuracy": 0.11961946487426758,
      "num_tokens": 2000202.0,
      "step": 1150
    },
    {
      "entropy": 6.645785331726074,
      "epoch": 0.08986578486675743,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499998380665456,
      "loss": 6.3866,
      "mean_token_accuracy": 0.11929098591208458,
      "num_tokens": 2009281.0,
      "step": 1155
    },
    {
      "entropy": 6.7169397354125975,
      "epoch": 0.090254814238475,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999982738069924,
      "loss": 6.4005,
      "mean_token_accuracy": 0.1197192907333374,
      "num_tokens": 2017806.0,
      "step": 1160
    },
    {
      "entropy": 6.742453861236572,
      "epoch": 0.09064384361019256,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999981635345382,
      "loss": 6.4894,
      "mean_token_accuracy": 0.11311474218964576,
      "num_tokens": 2027582.0,
      "step": 1165
    },
    {
      "entropy": 6.615994262695312,
      "epoch": 0.09103287298191014,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999980498480949,
      "loss": 6.3583,
      "mean_token_accuracy": 0.12255421206355095,
      "num_tokens": 2035427.0,
      "step": 1170
    },
    {
      "entropy": 6.8117693901062015,
      "epoch": 0.0914219023536277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999979327476644,
      "loss": 6.5496,
      "mean_token_accuracy": 0.1146990828216076,
      "num_tokens": 2044603.0,
      "step": 1175
    },
    {
      "entropy": 6.670216655731201,
      "epoch": 0.09181093172534527,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999978122332485,
      "loss": 6.3796,
      "mean_token_accuracy": 0.12255444154143333,
      "num_tokens": 2053272.0,
      "step": 1180
    },
    {
      "entropy": 6.7155945777893065,
      "epoch": 0.09219996109706283,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999976883048487,
      "loss": 6.3888,
      "mean_token_accuracy": 0.11783261522650719,
      "num_tokens": 2061677.0,
      "step": 1185
    },
    {
      "entropy": 6.645085716247559,
      "epoch": 0.0925889904687804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999975609624673,
      "loss": 6.4177,
      "mean_token_accuracy": 0.11735781580209732,
      "num_tokens": 2070282.0,
      "step": 1190
    },
    {
      "entropy": 6.775166654586792,
      "epoch": 0.09297801984049796,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999974302061059,
      "loss": 6.5046,
      "mean_token_accuracy": 0.11407285407185555,
      "num_tokens": 2078812.0,
      "step": 1195
    },
    {
      "entropy": 6.684069871902466,
      "epoch": 0.09336704921221552,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999972960357666,
      "loss": 6.3885,
      "mean_token_accuracy": 0.11487993374466896,
      "num_tokens": 2087020.0,
      "step": 1200
    },
    {
      "entropy": 6.642553663253784,
      "epoch": 0.09375607858393309,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999971584514516,
      "loss": 6.3448,
      "mean_token_accuracy": 0.12319881394505501,
      "num_tokens": 2095933.0,
      "step": 1205
    },
    {
      "entropy": 6.58814435005188,
      "epoch": 0.09414510795565065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999970174531627,
      "loss": 6.3534,
      "mean_token_accuracy": 0.1218513771891594,
      "num_tokens": 2105218.0,
      "step": 1210
    },
    {
      "entropy": 6.586796474456787,
      "epoch": 0.09453413732736822,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999968730409022,
      "loss": 6.2685,
      "mean_token_accuracy": 0.12166600152850152,
      "num_tokens": 2113808.0,
      "step": 1215
    },
    {
      "entropy": 6.612871360778809,
      "epoch": 0.09492316669908578,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999967252146723,
      "loss": 6.4393,
      "mean_token_accuracy": 0.11809361055493355,
      "num_tokens": 2123014.0,
      "step": 1220
    },
    {
      "entropy": 6.7797685146331785,
      "epoch": 0.09531219607080335,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999965739744752,
      "loss": 6.3594,
      "mean_token_accuracy": 0.1224447838962078,
      "num_tokens": 2131840.0,
      "step": 1225
    },
    {
      "entropy": 6.6803430080413815,
      "epoch": 0.09570122544252091,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999964193203132,
      "loss": 6.4559,
      "mean_token_accuracy": 0.11718669459223748,
      "num_tokens": 2140467.0,
      "step": 1230
    },
    {
      "entropy": 6.644154119491577,
      "epoch": 0.09609025481423847,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999962612521888,
      "loss": 6.402,
      "mean_token_accuracy": 0.12155328691005707,
      "num_tokens": 2149282.0,
      "step": 1235
    },
    {
      "entropy": 6.682088994979859,
      "epoch": 0.09647928418595604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999960997701041,
      "loss": 6.3943,
      "mean_token_accuracy": 0.1177689179778099,
      "num_tokens": 2157781.0,
      "step": 1240
    },
    {
      "entropy": 6.6188617706298825,
      "epoch": 0.0968683135576736,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999959348740618,
      "loss": 6.3784,
      "mean_token_accuracy": 0.12058958411216736,
      "num_tokens": 2166482.0,
      "step": 1245
    },
    {
      "entropy": 6.618022489547729,
      "epoch": 0.09725734292939117,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999957665640641,
      "loss": 6.391,
      "mean_token_accuracy": 0.11966572850942611,
      "num_tokens": 2174559.0,
      "step": 1250
    },
    {
      "entropy": 6.600797843933106,
      "epoch": 0.09764637230110873,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499995594840114,
      "loss": 6.2864,
      "mean_token_accuracy": 0.1250573568046093,
      "num_tokens": 2182374.0,
      "step": 1255
    },
    {
      "entropy": 6.596623373031616,
      "epoch": 0.0980354016728263,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999954197022138,
      "loss": 6.3828,
      "mean_token_accuracy": 0.12434251680970192,
      "num_tokens": 2191427.0,
      "step": 1260
    },
    {
      "entropy": 6.724891614913941,
      "epoch": 0.09842443104454386,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999952411503661,
      "loss": 6.4555,
      "mean_token_accuracy": 0.11615066304802894,
      "num_tokens": 2199698.0,
      "step": 1265
    },
    {
      "entropy": 6.649227571487427,
      "epoch": 0.09881346041626143,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999950591845738,
      "loss": 6.3025,
      "mean_token_accuracy": 0.12375782802700996,
      "num_tokens": 2208519.0,
      "step": 1270
    },
    {
      "entropy": 6.540418243408203,
      "epoch": 0.09920248978797899,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999948738048397,
      "loss": 6.3256,
      "mean_token_accuracy": 0.12646739184856415,
      "num_tokens": 2217453.0,
      "step": 1275
    },
    {
      "entropy": 6.652591800689697,
      "epoch": 0.09959151915969655,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999946850111663,
      "loss": 6.4342,
      "mean_token_accuracy": 0.12071752622723579,
      "num_tokens": 2225623.0,
      "step": 1280
    },
    {
      "entropy": 6.616247844696045,
      "epoch": 0.09998054853141412,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999944928035569,
      "loss": 6.3775,
      "mean_token_accuracy": 0.12225224003195763,
      "num_tokens": 2234519.0,
      "step": 1285
    },
    {
      "entropy": 6.634349060058594,
      "epoch": 0.10036957790313168,
      "grad_norm": 1.28125,
      "learning_rate": 0.000499994297182014,
      "loss": 6.3755,
      "mean_token_accuracy": 0.11769431307911873,
      "num_tokens": 2242994.0,
      "step": 1290
    },
    {
      "entropy": 6.680039644241333,
      "epoch": 0.10075860727484925,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999940981465408,
      "loss": 6.3857,
      "mean_token_accuracy": 0.12219717204570771,
      "num_tokens": 2251687.0,
      "step": 1295
    },
    {
      "entropy": 6.514635276794434,
      "epoch": 0.10114763664656681,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999938956971404,
      "loss": 6.3129,
      "mean_token_accuracy": 0.1261261597275734,
      "num_tokens": 2260751.0,
      "step": 1300
    },
    {
      "entropy": 6.620299577713013,
      "epoch": 0.10153666601828439,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999936898338156,
      "loss": 6.2472,
      "mean_token_accuracy": 0.13221826106309892,
      "num_tokens": 2269125.0,
      "step": 1305
    },
    {
      "entropy": 6.51522011756897,
      "epoch": 0.10192569539000194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999934805565698,
      "loss": 6.1888,
      "mean_token_accuracy": 0.125250905752182,
      "num_tokens": 2277098.0,
      "step": 1310
    },
    {
      "entropy": 6.529193973541259,
      "epoch": 0.10231472476171952,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499993267865406,
      "loss": 6.4692,
      "mean_token_accuracy": 0.1198497474193573,
      "num_tokens": 2285953.0,
      "step": 1315
    },
    {
      "entropy": 6.745606851577759,
      "epoch": 0.10270375413343708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999930517603275,
      "loss": 6.3768,
      "mean_token_accuracy": 0.1258055381476879,
      "num_tokens": 2294487.0,
      "step": 1320
    },
    {
      "entropy": 6.622673559188843,
      "epoch": 0.10309278350515463,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999928322413374,
      "loss": 6.4086,
      "mean_token_accuracy": 0.11929989755153655,
      "num_tokens": 2302915.0,
      "step": 1325
    },
    {
      "entropy": 6.561403083801269,
      "epoch": 0.1034818128768722,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999926093084394,
      "loss": 6.3663,
      "mean_token_accuracy": 0.11803061217069626,
      "num_tokens": 2310769.0,
      "step": 1330
    },
    {
      "entropy": 6.5140684127807615,
      "epoch": 0.10387084224858976,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999923829616365,
      "loss": 6.2494,
      "mean_token_accuracy": 0.12241962403059006,
      "num_tokens": 2319987.0,
      "step": 1335
    },
    {
      "entropy": 6.584002923965454,
      "epoch": 0.10425987162030734,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999921532009323,
      "loss": 6.3661,
      "mean_token_accuracy": 0.11921068355441093,
      "num_tokens": 2328964.0,
      "step": 1340
    },
    {
      "entropy": 6.654835939407349,
      "epoch": 0.1046489009920249,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999919200263304,
      "loss": 6.3291,
      "mean_token_accuracy": 0.1160627231001854,
      "num_tokens": 2337945.0,
      "step": 1345
    },
    {
      "entropy": 6.591136741638183,
      "epoch": 0.10503793036374247,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999916834378342,
      "loss": 6.2104,
      "mean_token_accuracy": 0.12765597626566888,
      "num_tokens": 2345989.0,
      "step": 1350
    },
    {
      "entropy": 6.478854656219482,
      "epoch": 0.10542695973546003,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999914434354472,
      "loss": 6.3382,
      "mean_token_accuracy": 0.12092916443943977,
      "num_tokens": 2353752.0,
      "step": 1355
    },
    {
      "entropy": 6.624578857421875,
      "epoch": 0.10581598910717759,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999912000191733,
      "loss": 6.3137,
      "mean_token_accuracy": 0.11746641173958779,
      "num_tokens": 2361961.0,
      "step": 1360
    },
    {
      "entropy": 6.488426542282104,
      "epoch": 0.10620501847889516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499990953189016,
      "loss": 6.2343,
      "mean_token_accuracy": 0.12489335536956787,
      "num_tokens": 2370148.0,
      "step": 1365
    },
    {
      "entropy": 6.567029905319214,
      "epoch": 0.10659404785061272,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999907029449791,
      "loss": 6.26,
      "mean_token_accuracy": 0.12806575298309325,
      "num_tokens": 2379105.0,
      "step": 1370
    },
    {
      "entropy": 6.575870656967163,
      "epoch": 0.10698307722233029,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999904492870663,
      "loss": 6.3367,
      "mean_token_accuracy": 0.1279653139412403,
      "num_tokens": 2387417.0,
      "step": 1375
    },
    {
      "entropy": 6.600449657440185,
      "epoch": 0.10737210659404785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999901922152816,
      "loss": 6.351,
      "mean_token_accuracy": 0.12834431305527688,
      "num_tokens": 2396163.0,
      "step": 1380
    },
    {
      "entropy": 6.425090503692627,
      "epoch": 0.10776113596576542,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999899317296288,
      "loss": 6.2181,
      "mean_token_accuracy": 0.1259994737803936,
      "num_tokens": 2404661.0,
      "step": 1385
    },
    {
      "entropy": 6.521142292022705,
      "epoch": 0.10815016533748298,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999896678301121,
      "loss": 6.3092,
      "mean_token_accuracy": 0.12282017320394516,
      "num_tokens": 2413459.0,
      "step": 1390
    },
    {
      "entropy": 6.661787652969361,
      "epoch": 0.10853919470920055,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999894005167352,
      "loss": 6.2619,
      "mean_token_accuracy": 0.1296503096818924,
      "num_tokens": 2421401.0,
      "step": 1395
    },
    {
      "entropy": 6.4945494651794435,
      "epoch": 0.10892822408091811,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499989129789502,
      "loss": 6.2816,
      "mean_token_accuracy": 0.1270440138876438,
      "num_tokens": 2429983.0,
      "step": 1400
    },
    {
      "entropy": 6.612727165222168,
      "epoch": 0.10931725345263567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999888556484172,
      "loss": 6.404,
      "mean_token_accuracy": 0.11439228877425194,
      "num_tokens": 2439206.0,
      "step": 1405
    },
    {
      "entropy": 6.542930793762207,
      "epoch": 0.10970628282435324,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999885780934845,
      "loss": 6.322,
      "mean_token_accuracy": 0.13015031218528747,
      "num_tokens": 2448801.0,
      "step": 1410
    },
    {
      "entropy": 6.549044942855835,
      "epoch": 0.1100953121960708,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999882971247081,
      "loss": 6.3615,
      "mean_token_accuracy": 0.12094596326351166,
      "num_tokens": 2458007.0,
      "step": 1415
    },
    {
      "entropy": 6.577499437332153,
      "epoch": 0.11048434156778837,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999880127420926,
      "loss": 6.2512,
      "mean_token_accuracy": 0.13192354440689086,
      "num_tokens": 2466574.0,
      "step": 1420
    },
    {
      "entropy": 6.569367551803589,
      "epoch": 0.11087337093950593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999877249456421,
      "loss": 6.3249,
      "mean_token_accuracy": 0.12187402471899986,
      "num_tokens": 2475263.0,
      "step": 1425
    },
    {
      "entropy": 6.3397541522979735,
      "epoch": 0.1112624003112235,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999874337353609,
      "loss": 6.0852,
      "mean_token_accuracy": 0.12592738643288612,
      "num_tokens": 2483738.0,
      "step": 1430
    },
    {
      "entropy": 6.452084875106811,
      "epoch": 0.11165142968294106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999871391112535,
      "loss": 6.1944,
      "mean_token_accuracy": 0.12284255996346474,
      "num_tokens": 2492937.0,
      "step": 1435
    },
    {
      "entropy": 6.443626451492309,
      "epoch": 0.11204045905465862,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999868410733244,
      "loss": 6.2673,
      "mean_token_accuracy": 0.12350725680589676,
      "num_tokens": 2500716.0,
      "step": 1440
    },
    {
      "entropy": 6.534166097640991,
      "epoch": 0.11242948842637619,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499986539621578,
      "loss": 6.2636,
      "mean_token_accuracy": 0.12070308104157448,
      "num_tokens": 2509385.0,
      "step": 1445
    },
    {
      "entropy": 6.586621427536011,
      "epoch": 0.11281851779809375,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999862347560191,
      "loss": 6.3272,
      "mean_token_accuracy": 0.11826682835817337,
      "num_tokens": 2518179.0,
      "step": 1450
    },
    {
      "entropy": 6.551055431365967,
      "epoch": 0.11320754716981132,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999859264766521,
      "loss": 6.2799,
      "mean_token_accuracy": 0.12288788110017776,
      "num_tokens": 2526644.0,
      "step": 1455
    },
    {
      "entropy": 6.4282551288604735,
      "epoch": 0.11359657654152888,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999856147834817,
      "loss": 6.1961,
      "mean_token_accuracy": 0.12372745499014855,
      "num_tokens": 2534998.0,
      "step": 1460
    },
    {
      "entropy": 6.5795485496521,
      "epoch": 0.11398560591324645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999852996765129,
      "loss": 6.2592,
      "mean_token_accuracy": 0.1275986485183239,
      "num_tokens": 2543562.0,
      "step": 1465
    },
    {
      "entropy": 6.49003233909607,
      "epoch": 0.11437463528496401,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999849811557502,
      "loss": 6.2683,
      "mean_token_accuracy": 0.12322987839579583,
      "num_tokens": 2551688.0,
      "step": 1470
    },
    {
      "entropy": 6.513979959487915,
      "epoch": 0.11476366465668159,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999846592211984,
      "loss": 6.3312,
      "mean_token_accuracy": 0.12700059562921523,
      "num_tokens": 2560227.0,
      "step": 1475
    },
    {
      "entropy": 6.532626295089722,
      "epoch": 0.11515269402839914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999843338728626,
      "loss": 6.2152,
      "mean_token_accuracy": 0.12785592749714852,
      "num_tokens": 2569245.0,
      "step": 1480
    },
    {
      "entropy": 6.480125904083252,
      "epoch": 0.1155417234001167,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999840051107476,
      "loss": 6.24,
      "mean_token_accuracy": 0.12533404007554055,
      "num_tokens": 2577706.0,
      "step": 1485
    },
    {
      "entropy": 6.5578601360321045,
      "epoch": 0.11593075277183428,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999836729348584,
      "loss": 6.274,
      "mean_token_accuracy": 0.12649679258465768,
      "num_tokens": 2586075.0,
      "step": 1490
    },
    {
      "entropy": 6.430877065658569,
      "epoch": 0.11631978214355183,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999833373452,
      "loss": 6.2685,
      "mean_token_accuracy": 0.12222606092691421,
      "num_tokens": 2594965.0,
      "step": 1495
    },
    {
      "entropy": 6.507078981399536,
      "epoch": 0.1167088115152694,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999829983417778,
      "loss": 6.1827,
      "mean_token_accuracy": 0.12774747908115386,
      "num_tokens": 2603355.0,
      "step": 1500
    },
    {
      "entropy": 6.502455472946167,
      "epoch": 0.11709784088698696,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999826559245965,
      "loss": 6.3264,
      "mean_token_accuracy": 0.12180719375610352,
      "num_tokens": 2611697.0,
      "step": 1505
    },
    {
      "entropy": 6.491758203506469,
      "epoch": 0.11748687025870454,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999823100936615,
      "loss": 6.2036,
      "mean_token_accuracy": 0.12712125554680825,
      "num_tokens": 2620821.0,
      "step": 1510
    },
    {
      "entropy": 6.50208306312561,
      "epoch": 0.1178758996304221,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999819608489781,
      "loss": 6.2029,
      "mean_token_accuracy": 0.12601106688380243,
      "num_tokens": 2628891.0,
      "step": 1515
    },
    {
      "entropy": 6.50620698928833,
      "epoch": 0.11826492900213967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999816081905515,
      "loss": 6.3238,
      "mean_token_accuracy": 0.11945099160075187,
      "num_tokens": 2638641.0,
      "step": 1520
    },
    {
      "entropy": 6.499932670593262,
      "epoch": 0.11865395837385723,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999812521183872,
      "loss": 6.2141,
      "mean_token_accuracy": 0.1259573705494404,
      "num_tokens": 2646898.0,
      "step": 1525
    },
    {
      "entropy": 6.5239215850830075,
      "epoch": 0.11904298774557479,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999808926324903,
      "loss": 6.2766,
      "mean_token_accuracy": 0.12060346677899361,
      "num_tokens": 2655404.0,
      "step": 1530
    },
    {
      "entropy": 6.430968427658081,
      "epoch": 0.11943201711729236,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999805297328665,
      "loss": 6.1457,
      "mean_token_accuracy": 0.1309056155383587,
      "num_tokens": 2663851.0,
      "step": 1535
    },
    {
      "entropy": 6.474033451080322,
      "epoch": 0.11982104648900992,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999801634195214,
      "loss": 6.2032,
      "mean_token_accuracy": 0.12700969725847244,
      "num_tokens": 2672231.0,
      "step": 1540
    },
    {
      "entropy": 6.383044147491455,
      "epoch": 0.12021007586072749,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999797936924603,
      "loss": 6.2097,
      "mean_token_accuracy": 0.12720894515514375,
      "num_tokens": 2680198.0,
      "step": 1545
    },
    {
      "entropy": 6.566014909744263,
      "epoch": 0.12059910523244505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499979420551689,
      "loss": 6.2435,
      "mean_token_accuracy": 0.12246377095580101,
      "num_tokens": 2689208.0,
      "step": 1550
    },
    {
      "entropy": 6.521243476867676,
      "epoch": 0.12098813460416262,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499979043997213,
      "loss": 6.2938,
      "mean_token_accuracy": 0.12263652533292771,
      "num_tokens": 2698116.0,
      "step": 1555
    },
    {
      "entropy": 6.459107732772827,
      "epoch": 0.12137716397588018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999786640290381,
      "loss": 6.2054,
      "mean_token_accuracy": 0.12052173987030983,
      "num_tokens": 2707046.0,
      "step": 1560
    },
    {
      "entropy": 6.418560600280761,
      "epoch": 0.12176619334759774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999782806471701,
      "loss": 6.177,
      "mean_token_accuracy": 0.12391525506973267,
      "num_tokens": 2714929.0,
      "step": 1565
    },
    {
      "entropy": 6.417961502075196,
      "epoch": 0.12215522271931531,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999778938516147,
      "loss": 6.2133,
      "mean_token_accuracy": 0.12548111602663994,
      "num_tokens": 2723436.0,
      "step": 1570
    },
    {
      "entropy": 6.4104386329650875,
      "epoch": 0.12254425209103287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499977503642378,
      "loss": 6.2541,
      "mean_token_accuracy": 0.12425186485052109,
      "num_tokens": 2731935.0,
      "step": 1575
    },
    {
      "entropy": 6.443300056457519,
      "epoch": 0.12293328146275044,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999771100194656,
      "loss": 6.0876,
      "mean_token_accuracy": 0.13300335481762887,
      "num_tokens": 2740307.0,
      "step": 1580
    },
    {
      "entropy": 6.338034725189209,
      "epoch": 0.123322310834468,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999767129828837,
      "loss": 6.1018,
      "mean_token_accuracy": 0.13122726380825042,
      "num_tokens": 2749077.0,
      "step": 1585
    },
    {
      "entropy": 6.4701416969299315,
      "epoch": 0.12371134020618557,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999763125326383,
      "loss": 6.2168,
      "mean_token_accuracy": 0.12340875789523124,
      "num_tokens": 2758133.0,
      "step": 1590
    },
    {
      "entropy": 6.40848298072815,
      "epoch": 0.12410036957790313,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999759086687355,
      "loss": 6.1445,
      "mean_token_accuracy": 0.13228374421596528,
      "num_tokens": 2766509.0,
      "step": 1595
    },
    {
      "entropy": 6.386842536926269,
      "epoch": 0.1244893989496207,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999755013911813,
      "loss": 6.3118,
      "mean_token_accuracy": 0.1265620172023773,
      "num_tokens": 2775908.0,
      "step": 1600
    },
    {
      "entropy": 6.4204363346099855,
      "epoch": 0.12487842832133826,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499975090699982,
      "loss": 6.0838,
      "mean_token_accuracy": 0.13560190349817275,
      "num_tokens": 2784803.0,
      "step": 1605
    },
    {
      "entropy": 6.483498287200928,
      "epoch": 0.12526745769305583,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999746765951438,
      "loss": 6.2464,
      "mean_token_accuracy": 0.1212728351354599,
      "num_tokens": 2793897.0,
      "step": 1610
    },
    {
      "entropy": 6.408759880065918,
      "epoch": 0.1256564870647734,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999742590766729,
      "loss": 6.0687,
      "mean_token_accuracy": 0.13336466401815414,
      "num_tokens": 2802652.0,
      "step": 1615
    },
    {
      "entropy": 6.33557162284851,
      "epoch": 0.12604551643649095,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999738381445757,
      "loss": 6.2091,
      "mean_token_accuracy": 0.12232308834791183,
      "num_tokens": 2811527.0,
      "step": 1620
    },
    {
      "entropy": 6.483191967010498,
      "epoch": 0.1264345458082085,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999734137988586,
      "loss": 6.0784,
      "mean_token_accuracy": 0.13252645283937453,
      "num_tokens": 2819410.0,
      "step": 1625
    },
    {
      "entropy": 6.221581315994262,
      "epoch": 0.1268235751799261,
      "grad_norm": 1.0,
      "learning_rate": 0.000499972986039528,
      "loss": 6.0676,
      "mean_token_accuracy": 0.13655331283807753,
      "num_tokens": 2828079.0,
      "step": 1630
    },
    {
      "entropy": 6.443491744995117,
      "epoch": 0.12721260455164365,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999725548665904,
      "loss": 6.186,
      "mean_token_accuracy": 0.1318517453968525,
      "num_tokens": 2836326.0,
      "step": 1635
    },
    {
      "entropy": 6.40787935256958,
      "epoch": 0.1276016339233612,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999721202800524,
      "loss": 6.2349,
      "mean_token_accuracy": 0.12275687456130982,
      "num_tokens": 2845271.0,
      "step": 1640
    },
    {
      "entropy": 6.344908809661865,
      "epoch": 0.12799066329507877,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999716822799206,
      "loss": 6.1049,
      "mean_token_accuracy": 0.13598963469266892,
      "num_tokens": 2853688.0,
      "step": 1645
    },
    {
      "entropy": 6.402085828781128,
      "epoch": 0.12837969266679633,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999712408662016,
      "loss": 6.1851,
      "mean_token_accuracy": 0.13216232731938363,
      "num_tokens": 2862261.0,
      "step": 1650
    },
    {
      "entropy": 6.434685897827149,
      "epoch": 0.12876872203851392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999707960389021,
      "loss": 6.215,
      "mean_token_accuracy": 0.12633103504776955,
      "num_tokens": 2871324.0,
      "step": 1655
    },
    {
      "entropy": 6.462427854537964,
      "epoch": 0.12915775141023147,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999703477980288,
      "loss": 6.2637,
      "mean_token_accuracy": 0.1204333670437336,
      "num_tokens": 2880268.0,
      "step": 1660
    },
    {
      "entropy": 6.463560485839844,
      "epoch": 0.12954678078194903,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999698961435885,
      "loss": 6.1692,
      "mean_token_accuracy": 0.12939711883664132,
      "num_tokens": 2888510.0,
      "step": 1665
    },
    {
      "entropy": 6.391593933105469,
      "epoch": 0.1299358101536666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999694410755882,
      "loss": 6.0645,
      "mean_token_accuracy": 0.13939797282218933,
      "num_tokens": 2897616.0,
      "step": 1670
    },
    {
      "entropy": 6.344762563705444,
      "epoch": 0.13032483952538418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999689825940347,
      "loss": 6.1536,
      "mean_token_accuracy": 0.13106709420681,
      "num_tokens": 2905703.0,
      "step": 1675
    },
    {
      "entropy": 6.3409374237060545,
      "epoch": 0.13071386889710174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999685206989349,
      "loss": 6.0402,
      "mean_token_accuracy": 0.1389915518462658,
      "num_tokens": 2914881.0,
      "step": 1680
    },
    {
      "entropy": 6.31886305809021,
      "epoch": 0.1311028982688193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999680553902959,
      "loss": 6.0988,
      "mean_token_accuracy": 0.13385162875056267,
      "num_tokens": 2923563.0,
      "step": 1685
    },
    {
      "entropy": 6.459726095199585,
      "epoch": 0.13149192764053685,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999675866681247,
      "loss": 6.3,
      "mean_token_accuracy": 0.12524155154824257,
      "num_tokens": 2932821.0,
      "step": 1690
    },
    {
      "entropy": 6.422298765182495,
      "epoch": 0.1318809570122544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999671145324286,
      "loss": 6.2041,
      "mean_token_accuracy": 0.12897164598107339,
      "num_tokens": 2940926.0,
      "step": 1695
    },
    {
      "entropy": 6.406879520416259,
      "epoch": 0.132269986383972,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999666389832144,
      "loss": 6.1352,
      "mean_token_accuracy": 0.13091752603650092,
      "num_tokens": 2950028.0,
      "step": 1700
    },
    {
      "entropy": 6.331121397018433,
      "epoch": 0.13265901575568956,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999661600204897,
      "loss": 6.158,
      "mean_token_accuracy": 0.13076577857136726,
      "num_tokens": 2957636.0,
      "step": 1705
    },
    {
      "entropy": 6.264277982711792,
      "epoch": 0.13304804512740712,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999656776442615,
      "loss": 5.9208,
      "mean_token_accuracy": 0.1367864117026329,
      "num_tokens": 2966544.0,
      "step": 1710
    },
    {
      "entropy": 6.367716217041016,
      "epoch": 0.13343707449912467,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999651918545372,
      "loss": 6.156,
      "mean_token_accuracy": 0.12581324428319932,
      "num_tokens": 2976236.0,
      "step": 1715
    },
    {
      "entropy": 6.343806362152099,
      "epoch": 0.13382610387084226,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999647026513243,
      "loss": 6.1055,
      "mean_token_accuracy": 0.1320023201406002,
      "num_tokens": 2984261.0,
      "step": 1720
    },
    {
      "entropy": 6.298451852798462,
      "epoch": 0.13421513324255982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049996421003463,
      "loss": 6.101,
      "mean_token_accuracy": 0.1286357559263706,
      "num_tokens": 2992970.0,
      "step": 1725
    },
    {
      "entropy": 6.399670839309692,
      "epoch": 0.13460416261427738,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999637140044619,
      "loss": 6.1143,
      "mean_token_accuracy": 0.13840305656194687,
      "num_tokens": 3001356.0,
      "step": 1730
    },
    {
      "entropy": 6.320483160018921,
      "epoch": 0.13499319198599494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999632145608275,
      "loss": 6.1338,
      "mean_token_accuracy": 0.13537682592868805,
      "num_tokens": 3009509.0,
      "step": 1735
    },
    {
      "entropy": 6.370671033859253,
      "epoch": 0.1353822213577125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999627117037343,
      "loss": 6.1434,
      "mean_token_accuracy": 0.12680430486798286,
      "num_tokens": 3017675.0,
      "step": 1740
    },
    {
      "entropy": 6.366966485977173,
      "epoch": 0.13577125072943008,
      "grad_norm": 1.203125,
      "learning_rate": 0.00049996220543319,
      "loss": 6.1149,
      "mean_token_accuracy": 0.1355073034763336,
      "num_tokens": 3025527.0,
      "step": 1745
    },
    {
      "entropy": 6.313861942291259,
      "epoch": 0.13616028010114764,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999616957492025,
      "loss": 6.0614,
      "mean_token_accuracy": 0.13035932257771493,
      "num_tokens": 3034251.0,
      "step": 1750
    },
    {
      "entropy": 6.291455030441284,
      "epoch": 0.1365493094728652,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999611826517793,
      "loss": 6.2936,
      "mean_token_accuracy": 0.12496695667505264,
      "num_tokens": 3043272.0,
      "step": 1755
    },
    {
      "entropy": 6.394519853591919,
      "epoch": 0.13693833884458276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999606661409281,
      "loss": 6.1376,
      "mean_token_accuracy": 0.13365355953574182,
      "num_tokens": 3051092.0,
      "step": 1760
    },
    {
      "entropy": 6.389160776138306,
      "epoch": 0.13732736821630034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999601462166569,
      "loss": 6.1093,
      "mean_token_accuracy": 0.13206425607204436,
      "num_tokens": 3060584.0,
      "step": 1765
    },
    {
      "entropy": 6.292361640930176,
      "epoch": 0.1377163975880179,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999596228789736,
      "loss": 6.112,
      "mean_token_accuracy": 0.13125555962324142,
      "num_tokens": 3069990.0,
      "step": 1770
    },
    {
      "entropy": 6.413235998153686,
      "epoch": 0.13810542695973546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999590961278859,
      "loss": 6.0901,
      "mean_token_accuracy": 0.1285365678369999,
      "num_tokens": 3078837.0,
      "step": 1775
    },
    {
      "entropy": 6.21497654914856,
      "epoch": 0.13849445633145302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999585659634021,
      "loss": 6.0786,
      "mean_token_accuracy": 0.1374272361397743,
      "num_tokens": 3087557.0,
      "step": 1780
    },
    {
      "entropy": 6.306288623809815,
      "epoch": 0.13888348570317058,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999580323855301,
      "loss": 6.086,
      "mean_token_accuracy": 0.13215372562408448,
      "num_tokens": 3096524.0,
      "step": 1785
    },
    {
      "entropy": 6.460636615753174,
      "epoch": 0.13927251507488816,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999574953942781,
      "loss": 6.1695,
      "mean_token_accuracy": 0.1295361652970314,
      "num_tokens": 3105271.0,
      "step": 1790
    },
    {
      "entropy": 6.278649234771729,
      "epoch": 0.13966154444660572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999569549896541,
      "loss": 6.0498,
      "mean_token_accuracy": 0.13072569519281388,
      "num_tokens": 3114479.0,
      "step": 1795
    },
    {
      "entropy": 6.313834571838379,
      "epoch": 0.14005057381832328,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999564111716665,
      "loss": 6.0575,
      "mean_token_accuracy": 0.13794643208384513,
      "num_tokens": 3123319.0,
      "step": 1800
    },
    {
      "entropy": 6.3173986911773685,
      "epoch": 0.14043960319004084,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999558639403232,
      "loss": 6.0822,
      "mean_token_accuracy": 0.13311919420957566,
      "num_tokens": 3132532.0,
      "step": 1805
    },
    {
      "entropy": 6.4103357791900635,
      "epoch": 0.1408286325617584,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999553132956328,
      "loss": 6.155,
      "mean_token_accuracy": 0.13053008466958999,
      "num_tokens": 3140798.0,
      "step": 1810
    },
    {
      "entropy": 6.330409145355224,
      "epoch": 0.14121766193347599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999547592376035,
      "loss": 6.1438,
      "mean_token_accuracy": 0.13290951102972032,
      "num_tokens": 3149607.0,
      "step": 1815
    },
    {
      "entropy": 6.251868629455567,
      "epoch": 0.14160669130519354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999542017662438,
      "loss": 5.9608,
      "mean_token_accuracy": 0.13571959659457206,
      "num_tokens": 3157537.0,
      "step": 1820
    },
    {
      "entropy": 6.296808433532715,
      "epoch": 0.1419957206769111,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999536408815623,
      "loss": 6.119,
      "mean_token_accuracy": 0.13132073506712913,
      "num_tokens": 3166351.0,
      "step": 1825
    },
    {
      "entropy": 6.25454888343811,
      "epoch": 0.14238475004862866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999530765835672,
      "loss": 5.9837,
      "mean_token_accuracy": 0.13595960065722465,
      "num_tokens": 3174690.0,
      "step": 1830
    },
    {
      "entropy": 6.26420259475708,
      "epoch": 0.14277377942034625,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999525088722673,
      "loss": 6.1311,
      "mean_token_accuracy": 0.12212195172905922,
      "num_tokens": 3183460.0,
      "step": 1835
    },
    {
      "entropy": 6.3232808113098145,
      "epoch": 0.1431628087920638,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999519377476711,
      "loss": 6.0604,
      "mean_token_accuracy": 0.13456608951091767,
      "num_tokens": 3192183.0,
      "step": 1840
    },
    {
      "entropy": 6.307611131668091,
      "epoch": 0.14355183816378136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999513632097872,
      "loss": 6.1451,
      "mean_token_accuracy": 0.13180193528532982,
      "num_tokens": 3200875.0,
      "step": 1845
    },
    {
      "entropy": 6.305696964263916,
      "epoch": 0.14394086753549892,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999507852586245,
      "loss": 6.0116,
      "mean_token_accuracy": 0.13570416271686553,
      "num_tokens": 3209797.0,
      "step": 1850
    },
    {
      "entropy": 6.369657182693482,
      "epoch": 0.14432989690721648,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999502038941917,
      "loss": 6.1304,
      "mean_token_accuracy": 0.12988198101520537,
      "num_tokens": 3218612.0,
      "step": 1855
    },
    {
      "entropy": 6.262934398651123,
      "epoch": 0.14471892627893407,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999496191164975,
      "loss": 6.0438,
      "mean_token_accuracy": 0.13271042779088021,
      "num_tokens": 3227364.0,
      "step": 1860
    },
    {
      "entropy": 6.2206947803497314,
      "epoch": 0.14510795565065163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499949030925551,
      "loss": 6.0766,
      "mean_token_accuracy": 0.13223439678549767,
      "num_tokens": 3236448.0,
      "step": 1865
    },
    {
      "entropy": 6.286829996109009,
      "epoch": 0.14549698502236919,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999484393213609,
      "loss": 6.0017,
      "mean_token_accuracy": 0.1383517123758793,
      "num_tokens": 3244957.0,
      "step": 1870
    },
    {
      "entropy": 6.170301628112793,
      "epoch": 0.14588601439408674,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999478443039363,
      "loss": 6.0811,
      "mean_token_accuracy": 0.12760961875319482,
      "num_tokens": 3253250.0,
      "step": 1875
    },
    {
      "entropy": 6.355124139785767,
      "epoch": 0.14627504376580433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999472458732862,
      "loss": 5.9909,
      "mean_token_accuracy": 0.13991812169551848,
      "num_tokens": 3261455.0,
      "step": 1880
    },
    {
      "entropy": 6.20726432800293,
      "epoch": 0.1466640731375219,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999466440294198,
      "loss": 6.0483,
      "mean_token_accuracy": 0.13473452106118203,
      "num_tokens": 3270108.0,
      "step": 1885
    },
    {
      "entropy": 6.375078630447388,
      "epoch": 0.14705310250923945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499946038772346,
      "loss": 6.1673,
      "mean_token_accuracy": 0.13430963680148125,
      "num_tokens": 3278425.0,
      "step": 1890
    },
    {
      "entropy": 6.323588514328003,
      "epoch": 0.147442131880957,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999454301020741,
      "loss": 6.0903,
      "mean_token_accuracy": 0.13201274499297141,
      "num_tokens": 3288358.0,
      "step": 1895
    },
    {
      "entropy": 6.309664011001587,
      "epoch": 0.14783116125267456,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999448180186133,
      "loss": 6.0894,
      "mean_token_accuracy": 0.13479083627462388,
      "num_tokens": 3297199.0,
      "step": 1900
    },
    {
      "entropy": 6.2201214790344235,
      "epoch": 0.14822019062439215,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499944202521973,
      "loss": 6.0839,
      "mean_token_accuracy": 0.13048910349607468,
      "num_tokens": 3306350.0,
      "step": 1905
    },
    {
      "entropy": 6.295196104049682,
      "epoch": 0.1486092199961097,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999435836121624,
      "loss": 5.9272,
      "mean_token_accuracy": 0.1357748366892338,
      "num_tokens": 3314921.0,
      "step": 1910
    },
    {
      "entropy": 6.121889209747314,
      "epoch": 0.14899824936782727,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499942961289191,
      "loss": 6.0676,
      "mean_token_accuracy": 0.13817773759365082,
      "num_tokens": 3323324.0,
      "step": 1915
    },
    {
      "entropy": 6.420400953292846,
      "epoch": 0.14938727873954483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999423355530683,
      "loss": 6.1954,
      "mean_token_accuracy": 0.12781462743878363,
      "num_tokens": 3331614.0,
      "step": 1920
    },
    {
      "entropy": 6.316415214538575,
      "epoch": 0.1497763081112624,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999417064038035,
      "loss": 6.0304,
      "mean_token_accuracy": 0.1352904126048088,
      "num_tokens": 3339573.0,
      "step": 1925
    },
    {
      "entropy": 6.221908378601074,
      "epoch": 0.15016533748297997,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999410738414065,
      "loss": 6.0006,
      "mean_token_accuracy": 0.1410306826233864,
      "num_tokens": 3347775.0,
      "step": 1930
    },
    {
      "entropy": 6.257770204544068,
      "epoch": 0.15055436685469753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999404378658866,
      "loss": 5.9657,
      "mean_token_accuracy": 0.13755407258868219,
      "num_tokens": 3356947.0,
      "step": 1935
    },
    {
      "entropy": 6.257419776916504,
      "epoch": 0.1509433962264151,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999397984772536,
      "loss": 6.0561,
      "mean_token_accuracy": 0.13719375282526017,
      "num_tokens": 3365420.0,
      "step": 1940
    },
    {
      "entropy": 6.31306824684143,
      "epoch": 0.15133242559813265,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999391556755172,
      "loss": 6.1417,
      "mean_token_accuracy": 0.13246554136276245,
      "num_tokens": 3373483.0,
      "step": 1945
    },
    {
      "entropy": 6.326482582092285,
      "epoch": 0.15172145496985023,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999385094606872,
      "loss": 6.1331,
      "mean_token_accuracy": 0.1271201863884926,
      "num_tokens": 3382557.0,
      "step": 1950
    },
    {
      "entropy": 6.310076713562012,
      "epoch": 0.1521104843415678,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999378598327733,
      "loss": 6.081,
      "mean_token_accuracy": 0.12705598026514053,
      "num_tokens": 3392079.0,
      "step": 1955
    },
    {
      "entropy": 6.259531545639038,
      "epoch": 0.15249951371328535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999372067917854,
      "loss": 6.1412,
      "mean_token_accuracy": 0.12986933141946794,
      "num_tokens": 3401070.0,
      "step": 1960
    },
    {
      "entropy": 6.24435453414917,
      "epoch": 0.1528885430850029,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999365503377335,
      "loss": 6.0231,
      "mean_token_accuracy": 0.13275498896837234,
      "num_tokens": 3409921.0,
      "step": 1965
    },
    {
      "entropy": 6.3059319972991945,
      "epoch": 0.1532775724567205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999358904706273,
      "loss": 6.0424,
      "mean_token_accuracy": 0.13057347238063813,
      "num_tokens": 3419149.0,
      "step": 1970
    },
    {
      "entropy": 6.208243465423584,
      "epoch": 0.15366660182843805,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999352271904771,
      "loss": 5.9333,
      "mean_token_accuracy": 0.14223069027066232,
      "num_tokens": 3427435.0,
      "step": 1975
    },
    {
      "entropy": 6.283069849014282,
      "epoch": 0.1540556312001556,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999345604972928,
      "loss": 6.1375,
      "mean_token_accuracy": 0.1275772549211979,
      "num_tokens": 3436377.0,
      "step": 1980
    },
    {
      "entropy": 6.239132833480835,
      "epoch": 0.15444466057187317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999338903910846,
      "loss": 6.0122,
      "mean_token_accuracy": 0.13495821878314018,
      "num_tokens": 3444997.0,
      "step": 1985
    },
    {
      "entropy": 6.343378400802612,
      "epoch": 0.15483368994359073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999332168718626,
      "loss": 6.0894,
      "mean_token_accuracy": 0.1357529453933239,
      "num_tokens": 3454254.0,
      "step": 1990
    },
    {
      "entropy": 6.249056434631347,
      "epoch": 0.15522271931530832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999325399396371,
      "loss": 6.0377,
      "mean_token_accuracy": 0.1396930918097496,
      "num_tokens": 3462853.0,
      "step": 1995
    },
    {
      "entropy": 6.227156352996826,
      "epoch": 0.15561174868702587,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999318595944183,
      "loss": 5.9889,
      "mean_token_accuracy": 0.13459807708859445,
      "num_tokens": 3471288.0,
      "step": 2000
    },
    {
      "entropy": 6.149478435516357,
      "epoch": 0.15600077805874343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999311758362166,
      "loss": 5.8852,
      "mean_token_accuracy": 0.13669946119189264,
      "num_tokens": 3479479.0,
      "step": 2005
    },
    {
      "entropy": 6.195598793029785,
      "epoch": 0.156389807430461,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999304886650422,
      "loss": 5.9151,
      "mean_token_accuracy": 0.1384771704673767,
      "num_tokens": 3488621.0,
      "step": 2010
    },
    {
      "entropy": 6.179077291488648,
      "epoch": 0.15677883680217858,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999297980809058,
      "loss": 5.9768,
      "mean_token_accuracy": 0.1390062913298607,
      "num_tokens": 3497792.0,
      "step": 2015
    },
    {
      "entropy": 6.147572040557861,
      "epoch": 0.15716786617389614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999291040838176,
      "loss": 5.9844,
      "mean_token_accuracy": 0.13780189007520677,
      "num_tokens": 3506548.0,
      "step": 2020
    },
    {
      "entropy": 6.185660696029663,
      "epoch": 0.1575568955456137,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999284066737884,
      "loss": 5.973,
      "mean_token_accuracy": 0.14213240072131156,
      "num_tokens": 3515086.0,
      "step": 2025
    },
    {
      "entropy": 6.301659440994262,
      "epoch": 0.15794592491733125,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999277058508285,
      "loss": 6.0185,
      "mean_token_accuracy": 0.13498132154345513,
      "num_tokens": 3523985.0,
      "step": 2030
    },
    {
      "entropy": 6.245706176757812,
      "epoch": 0.1583349542890488,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999270016149486,
      "loss": 6.0751,
      "mean_token_accuracy": 0.13686039298772812,
      "num_tokens": 3532575.0,
      "step": 2035
    },
    {
      "entropy": 6.304662990570068,
      "epoch": 0.1587239836607664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999262939661596,
      "loss": 6.0697,
      "mean_token_accuracy": 0.13059490993618966,
      "num_tokens": 3541558.0,
      "step": 2040
    },
    {
      "entropy": 6.182402086257935,
      "epoch": 0.15911301303248396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999255829044721,
      "loss": 5.9798,
      "mean_token_accuracy": 0.13890642076730728,
      "num_tokens": 3549697.0,
      "step": 2045
    },
    {
      "entropy": 6.313886213302612,
      "epoch": 0.15950204240420152,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999248684298968,
      "loss": 6.1551,
      "mean_token_accuracy": 0.12896221354603768,
      "num_tokens": 3558673.0,
      "step": 2050
    },
    {
      "entropy": 6.274616432189942,
      "epoch": 0.15989107177591907,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999241505424447,
      "loss": 6.0047,
      "mean_token_accuracy": 0.13846511095762254,
      "num_tokens": 3567176.0,
      "step": 2055
    },
    {
      "entropy": 6.335512018203735,
      "epoch": 0.16028010114763663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999234292421265,
      "loss": 6.1437,
      "mean_token_accuracy": 0.13223661482334137,
      "num_tokens": 3576020.0,
      "step": 2060
    },
    {
      "entropy": 6.19822096824646,
      "epoch": 0.16066913051935422,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999227045289535,
      "loss": 5.9723,
      "mean_token_accuracy": 0.13607832789421082,
      "num_tokens": 3584705.0,
      "step": 2065
    },
    {
      "entropy": 6.279978084564209,
      "epoch": 0.16105815989107178,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999219764029363,
      "loss": 6.1605,
      "mean_token_accuracy": 0.1307303138077259,
      "num_tokens": 3593339.0,
      "step": 2070
    },
    {
      "entropy": 6.261270904541016,
      "epoch": 0.16144718926278934,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999212448640861,
      "loss": 5.8723,
      "mean_token_accuracy": 0.14517469108104705,
      "num_tokens": 3601206.0,
      "step": 2075
    },
    {
      "entropy": 6.079396677017212,
      "epoch": 0.1618362186345069,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999205099124141,
      "loss": 5.9946,
      "mean_token_accuracy": 0.13344853073358537,
      "num_tokens": 3610371.0,
      "step": 2080
    },
    {
      "entropy": 6.325978946685791,
      "epoch": 0.16222524800622448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999197715479313,
      "loss": 6.0037,
      "mean_token_accuracy": 0.1363994300365448,
      "num_tokens": 3618834.0,
      "step": 2085
    },
    {
      "entropy": 6.110661888122559,
      "epoch": 0.16261427737794204,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999190297706489,
      "loss": 6.041,
      "mean_token_accuracy": 0.138499366492033,
      "num_tokens": 3627426.0,
      "step": 2090
    },
    {
      "entropy": 6.259611415863037,
      "epoch": 0.1630033067496596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999182845805783,
      "loss": 6.0427,
      "mean_token_accuracy": 0.13478516712784766,
      "num_tokens": 3636020.0,
      "step": 2095
    },
    {
      "entropy": 6.348205089569092,
      "epoch": 0.16339233612137716,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999175359777306,
      "loss": 5.9891,
      "mean_token_accuracy": 0.1379244014620781,
      "num_tokens": 3644419.0,
      "step": 2100
    },
    {
      "entropy": 6.189280939102173,
      "epoch": 0.16378136549309472,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999167839621174,
      "loss": 6.0854,
      "mean_token_accuracy": 0.12851483151316642,
      "num_tokens": 3653223.0,
      "step": 2105
    },
    {
      "entropy": 6.301689481735229,
      "epoch": 0.1641703948648123,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999160285337501,
      "loss": 6.1,
      "mean_token_accuracy": 0.1289041444659233,
      "num_tokens": 3662409.0,
      "step": 2110
    },
    {
      "entropy": 6.2754106521606445,
      "epoch": 0.16455942423652986,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999152696926399,
      "loss": 6.0583,
      "mean_token_accuracy": 0.13341646790504455,
      "num_tokens": 3670635.0,
      "step": 2115
    },
    {
      "entropy": 6.122903490066529,
      "epoch": 0.16494845360824742,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999145074387985,
      "loss": 5.9886,
      "mean_token_accuracy": 0.14272082224488258,
      "num_tokens": 3679303.0,
      "step": 2120
    },
    {
      "entropy": 6.311637830734253,
      "epoch": 0.16533748297996498,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999137417722374,
      "loss": 6.1483,
      "mean_token_accuracy": 0.13292766511440277,
      "num_tokens": 3689299.0,
      "step": 2125
    },
    {
      "entropy": 6.2762187957763675,
      "epoch": 0.16572651235168256,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999129726929684,
      "loss": 6.0652,
      "mean_token_accuracy": 0.13817578107118605,
      "num_tokens": 3697843.0,
      "step": 2130
    },
    {
      "entropy": 6.290596771240234,
      "epoch": 0.16611554172340012,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999122002010029,
      "loss": 5.9937,
      "mean_token_accuracy": 0.1410035215318203,
      "num_tokens": 3705876.0,
      "step": 2135
    },
    {
      "entropy": 6.264617252349853,
      "epoch": 0.16650457109511768,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999114242963527,
      "loss": 6.1131,
      "mean_token_accuracy": 0.12808321714401244,
      "num_tokens": 3714835.0,
      "step": 2140
    },
    {
      "entropy": 6.201760530471802,
      "epoch": 0.16689360046683524,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999106449790298,
      "loss": 5.9684,
      "mean_token_accuracy": 0.13607337027788163,
      "num_tokens": 3723421.0,
      "step": 2145
    },
    {
      "entropy": 6.197750616073608,
      "epoch": 0.1672826298385528,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999098622490458,
      "loss": 5.999,
      "mean_token_accuracy": 0.13562173172831535,
      "num_tokens": 3731590.0,
      "step": 2150
    },
    {
      "entropy": 6.155825185775757,
      "epoch": 0.16767165921027039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999090761064125,
      "loss": 5.9245,
      "mean_token_accuracy": 0.13581221401691437,
      "num_tokens": 3740700.0,
      "step": 2155
    },
    {
      "entropy": 6.188067531585693,
      "epoch": 0.16806068858198794,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499908286551142,
      "loss": 5.9976,
      "mean_token_accuracy": 0.14268267750740052,
      "num_tokens": 3749516.0,
      "step": 2160
    },
    {
      "entropy": 6.181734943389893,
      "epoch": 0.1684497179537055,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999074935832463,
      "loss": 5.9116,
      "mean_token_accuracy": 0.1446475252509117,
      "num_tokens": 3757696.0,
      "step": 2165
    },
    {
      "entropy": 6.187936496734619,
      "epoch": 0.16883874732542306,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999066972027373,
      "loss": 6.0052,
      "mean_token_accuracy": 0.13505456671118737,
      "num_tokens": 3765786.0,
      "step": 2170
    },
    {
      "entropy": 6.182317399978638,
      "epoch": 0.16922777669714065,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999058974096271,
      "loss": 5.9076,
      "mean_token_accuracy": 0.14293930530548096,
      "num_tokens": 3773265.0,
      "step": 2175
    },
    {
      "entropy": 6.148977756500244,
      "epoch": 0.1696168060688582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499905094203928,
      "loss": 6.029,
      "mean_token_accuracy": 0.13410733118653298,
      "num_tokens": 3782402.0,
      "step": 2180
    },
    {
      "entropy": 6.245903635025025,
      "epoch": 0.17000583544057576,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499904287585652,
      "loss": 6.0971,
      "mean_token_accuracy": 0.12724517583847045,
      "num_tokens": 3791901.0,
      "step": 2185
    },
    {
      "entropy": 6.184202718734741,
      "epoch": 0.17039486481229332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999034775548113,
      "loss": 5.9206,
      "mean_token_accuracy": 0.14507496878504753,
      "num_tokens": 3799872.0,
      "step": 2190
    },
    {
      "entropy": 6.159626245498657,
      "epoch": 0.17078389418401088,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999026641114185,
      "loss": 5.858,
      "mean_token_accuracy": 0.14604728296399117,
      "num_tokens": 3808949.0,
      "step": 2195
    },
    {
      "entropy": 6.170258522033691,
      "epoch": 0.17117292355572847,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999018472554857,
      "loss": 5.9815,
      "mean_token_accuracy": 0.13522427454590796,
      "num_tokens": 3817429.0,
      "step": 2200
    },
    {
      "entropy": 6.33716516494751,
      "epoch": 0.17156195292744603,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999010269870253,
      "loss": 6.2642,
      "mean_token_accuracy": 0.12795186191797256,
      "num_tokens": 3827101.0,
      "step": 2205
    },
    {
      "entropy": 6.283813047409057,
      "epoch": 0.17195098229916359,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999002033060498,
      "loss": 6.0133,
      "mean_token_accuracy": 0.1351143278181553,
      "num_tokens": 3836147.0,
      "step": 2210
    },
    {
      "entropy": 6.136712694168091,
      "epoch": 0.17234001167088114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998993762125716,
      "loss": 6.0013,
      "mean_token_accuracy": 0.1369711324572563,
      "num_tokens": 3845300.0,
      "step": 2215
    },
    {
      "entropy": 6.282448244094849,
      "epoch": 0.17272904104259873,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998985457066034,
      "loss": 5.9966,
      "mean_token_accuracy": 0.1379320077598095,
      "num_tokens": 3853620.0,
      "step": 2220
    },
    {
      "entropy": 6.123518180847168,
      "epoch": 0.1731180704143163,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998977117881576,
      "loss": 5.9928,
      "mean_token_accuracy": 0.1383661672472954,
      "num_tokens": 3861734.0,
      "step": 2225
    },
    {
      "entropy": 6.208690977096557,
      "epoch": 0.17350709978603385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998968744572472,
      "loss": 6.0124,
      "mean_token_accuracy": 0.13703416362404824,
      "num_tokens": 3870753.0,
      "step": 2230
    },
    {
      "entropy": 6.165704917907715,
      "epoch": 0.1738961291577514,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998960337138845,
      "loss": 5.8932,
      "mean_token_accuracy": 0.14122473895549775,
      "num_tokens": 3879382.0,
      "step": 2235
    },
    {
      "entropy": 6.184034299850464,
      "epoch": 0.17428515852946896,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998951895580826,
      "loss": 6.065,
      "mean_token_accuracy": 0.1349945433437824,
      "num_tokens": 3888417.0,
      "step": 2240
    },
    {
      "entropy": 6.127526903152466,
      "epoch": 0.17467418790118655,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998943419898542,
      "loss": 5.9209,
      "mean_token_accuracy": 0.13626997768878937,
      "num_tokens": 3897083.0,
      "step": 2245
    },
    {
      "entropy": 6.1350280284881595,
      "epoch": 0.1750632172729041,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998934910092119,
      "loss": 5.8862,
      "mean_token_accuracy": 0.1451417922973633,
      "num_tokens": 3905564.0,
      "step": 2250
    },
    {
      "entropy": 6.2061069965362545,
      "epoch": 0.17545224664462167,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998926366161689,
      "loss": 6.0361,
      "mean_token_accuracy": 0.1296758010983467,
      "num_tokens": 3914896.0,
      "step": 2255
    },
    {
      "entropy": 6.1870911598205565,
      "epoch": 0.17584127601633923,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998917788107383,
      "loss": 6.0419,
      "mean_token_accuracy": 0.14005873054265977,
      "num_tokens": 3924151.0,
      "step": 2260
    },
    {
      "entropy": 6.233715152740478,
      "epoch": 0.17623030538805678,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998909175929326,
      "loss": 5.9814,
      "mean_token_accuracy": 0.14144271165132521,
      "num_tokens": 3933163.0,
      "step": 2265
    },
    {
      "entropy": 6.163707447052002,
      "epoch": 0.17661933475977437,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998900529627653,
      "loss": 5.9032,
      "mean_token_accuracy": 0.1486126258969307,
      "num_tokens": 3941120.0,
      "step": 2270
    },
    {
      "entropy": 6.058203554153442,
      "epoch": 0.17700836413149193,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998891849202495,
      "loss": 5.7958,
      "mean_token_accuracy": 0.14378746151924132,
      "num_tokens": 3949896.0,
      "step": 2275
    },
    {
      "entropy": 6.195409440994263,
      "epoch": 0.1773973935032095,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499888313465398,
      "loss": 5.951,
      "mean_token_accuracy": 0.13810205608606338,
      "num_tokens": 3958794.0,
      "step": 2280
    },
    {
      "entropy": 6.186099672317505,
      "epoch": 0.17778642287492705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998874385982247,
      "loss": 6.0078,
      "mean_token_accuracy": 0.13773410096764566,
      "num_tokens": 3967267.0,
      "step": 2285
    },
    {
      "entropy": 6.139815187454223,
      "epoch": 0.17817545224664463,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998865603187421,
      "loss": 5.9249,
      "mean_token_accuracy": 0.13793863877654075,
      "num_tokens": 3976536.0,
      "step": 2290
    },
    {
      "entropy": 6.24169807434082,
      "epoch": 0.1785644816183622,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499885678626964,
      "loss": 5.9947,
      "mean_token_accuracy": 0.1388470858335495,
      "num_tokens": 3986006.0,
      "step": 2295
    },
    {
      "entropy": 6.077307844161988,
      "epoch": 0.17895351099007975,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998847935229038,
      "loss": 5.9394,
      "mean_token_accuracy": 0.1349871888756752,
      "num_tokens": 3994478.0,
      "step": 2300
    },
    {
      "entropy": 6.16282639503479,
      "epoch": 0.1793425403617973,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998839050065746,
      "loss": 5.9792,
      "mean_token_accuracy": 0.13701414689421654,
      "num_tokens": 4003190.0,
      "step": 2305
    },
    {
      "entropy": 6.2592230319976805,
      "epoch": 0.17973156973351487,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998830130779902,
      "loss": 5.9066,
      "mean_token_accuracy": 0.13989242389798165,
      "num_tokens": 4011886.0,
      "step": 2310
    },
    {
      "entropy": 6.044191455841064,
      "epoch": 0.18012059910523245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499882117737164,
      "loss": 5.8027,
      "mean_token_accuracy": 0.14078635722398758,
      "num_tokens": 4020792.0,
      "step": 2315
    },
    {
      "entropy": 6.006427335739136,
      "epoch": 0.18050962847695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998812189841096,
      "loss": 5.7981,
      "mean_token_accuracy": 0.1413320556282997,
      "num_tokens": 4029235.0,
      "step": 2320
    },
    {
      "entropy": 6.115646028518677,
      "epoch": 0.18089865784866757,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998803168188407,
      "loss": 6.0277,
      "mean_token_accuracy": 0.14016879796981813,
      "num_tokens": 4037341.0,
      "step": 2325
    },
    {
      "entropy": 6.140260648727417,
      "epoch": 0.18128768722038513,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998794112413708,
      "loss": 5.9019,
      "mean_token_accuracy": 0.13752360120415688,
      "num_tokens": 4046324.0,
      "step": 2330
    },
    {
      "entropy": 6.080597591400147,
      "epoch": 0.18167671659210272,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998785022517137,
      "loss": 5.9663,
      "mean_token_accuracy": 0.1393540196120739,
      "num_tokens": 4054579.0,
      "step": 2335
    },
    {
      "entropy": 6.197992992401123,
      "epoch": 0.18206574596382027,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998775898498835,
      "loss": 5.924,
      "mean_token_accuracy": 0.13897303566336633,
      "num_tokens": 4063087.0,
      "step": 2340
    },
    {
      "entropy": 6.189882612228393,
      "epoch": 0.18245477533553783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998766740358936,
      "loss": 5.9942,
      "mean_token_accuracy": 0.13766855299472808,
      "num_tokens": 4072309.0,
      "step": 2345
    },
    {
      "entropy": 6.12577452659607,
      "epoch": 0.1828438047072554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998757548097582,
      "loss": 6.045,
      "mean_token_accuracy": 0.1355599895119667,
      "num_tokens": 4081017.0,
      "step": 2350
    },
    {
      "entropy": 6.221224927902222,
      "epoch": 0.18323283407897295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998748321714911,
      "loss": 6.0571,
      "mean_token_accuracy": 0.13211404532194138,
      "num_tokens": 4089799.0,
      "step": 2355
    },
    {
      "entropy": 6.166183710098267,
      "epoch": 0.18362186345069054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998739061211065,
      "loss": 5.9713,
      "mean_token_accuracy": 0.14161632731556892,
      "num_tokens": 4097948.0,
      "step": 2360
    },
    {
      "entropy": 6.114394664764404,
      "epoch": 0.1840108928224081,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998729766586181,
      "loss": 5.8536,
      "mean_token_accuracy": 0.14067127406597138,
      "num_tokens": 4107050.0,
      "step": 2365
    },
    {
      "entropy": 6.107758188247681,
      "epoch": 0.18439992219412565,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998720437840403,
      "loss": 5.8672,
      "mean_token_accuracy": 0.13855508863925933,
      "num_tokens": 4115547.0,
      "step": 2370
    },
    {
      "entropy": 6.124674224853516,
      "epoch": 0.1847889515658432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998711074973871,
      "loss": 5.8706,
      "mean_token_accuracy": 0.14566295892000197,
      "num_tokens": 4124047.0,
      "step": 2375
    },
    {
      "entropy": 6.017377805709839,
      "epoch": 0.1851779809375608,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998701677986728,
      "loss": 5.8272,
      "mean_token_accuracy": 0.14891343787312508,
      "num_tokens": 4132005.0,
      "step": 2380
    },
    {
      "entropy": 6.186097240447998,
      "epoch": 0.18556701030927836,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998692246879116,
      "loss": 6.0144,
      "mean_token_accuracy": 0.13196168541908265,
      "num_tokens": 4140199.0,
      "step": 2385
    },
    {
      "entropy": 6.144238615036011,
      "epoch": 0.18595603968099592,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998682781651178,
      "loss": 5.9401,
      "mean_token_accuracy": 0.14200834929943085,
      "num_tokens": 4149353.0,
      "step": 2390
    },
    {
      "entropy": 6.03447151184082,
      "epoch": 0.18634506905271347,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998673282303059,
      "loss": 5.9088,
      "mean_token_accuracy": 0.14188055396080018,
      "num_tokens": 4158531.0,
      "step": 2395
    },
    {
      "entropy": 6.000899457931519,
      "epoch": 0.18673409842443103,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998663748834901,
      "loss": 5.7074,
      "mean_token_accuracy": 0.15226681232452394,
      "num_tokens": 4166743.0,
      "step": 2400
    },
    {
      "entropy": 6.190967559814453,
      "epoch": 0.18712312779614862,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998654181246851,
      "loss": 6.057,
      "mean_token_accuracy": 0.13685035333037376,
      "num_tokens": 4175416.0,
      "step": 2405
    },
    {
      "entropy": 6.181151866912842,
      "epoch": 0.18751215716786618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998644579539052,
      "loss": 5.9812,
      "mean_token_accuracy": 0.13951558768749237,
      "num_tokens": 4185324.0,
      "step": 2410
    },
    {
      "entropy": 6.096161127090454,
      "epoch": 0.18790118653958374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499863494371165,
      "loss": 5.9178,
      "mean_token_accuracy": 0.1447077937424183,
      "num_tokens": 4193361.0,
      "step": 2415
    },
    {
      "entropy": 6.187608957290649,
      "epoch": 0.1882902159113013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998625273764793,
      "loss": 5.9679,
      "mean_token_accuracy": 0.13389892131090164,
      "num_tokens": 4202553.0,
      "step": 2420
    },
    {
      "entropy": 6.0965659618377686,
      "epoch": 0.18867924528301888,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998615569698626,
      "loss": 5.9039,
      "mean_token_accuracy": 0.14300661832094191,
      "num_tokens": 4211585.0,
      "step": 2425
    },
    {
      "entropy": 6.0760973453521725,
      "epoch": 0.18906827465473644,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998605831513296,
      "loss": 5.9462,
      "mean_token_accuracy": 0.13661156445741654,
      "num_tokens": 4220829.0,
      "step": 2430
    },
    {
      "entropy": 6.0929618835449215,
      "epoch": 0.189457304026454,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998596059208953,
      "loss": 5.8543,
      "mean_token_accuracy": 0.14136664643883706,
      "num_tokens": 4228658.0,
      "step": 2435
    },
    {
      "entropy": 6.181618976593017,
      "epoch": 0.18984633339817156,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998586252785743,
      "loss": 5.9419,
      "mean_token_accuracy": 0.14073347449302673,
      "num_tokens": 4237650.0,
      "step": 2440
    },
    {
      "entropy": 6.056187391281128,
      "epoch": 0.19023536276988912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998576412243816,
      "loss": 5.9193,
      "mean_token_accuracy": 0.1453680008649826,
      "num_tokens": 4246701.0,
      "step": 2445
    },
    {
      "entropy": 6.117009496688842,
      "epoch": 0.1906243921416067,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998566537583321,
      "loss": 5.8866,
      "mean_token_accuracy": 0.14391229003667833,
      "num_tokens": 4254541.0,
      "step": 2450
    },
    {
      "entropy": 6.13186469078064,
      "epoch": 0.19101342151332426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998556628804408,
      "loss": 5.9241,
      "mean_token_accuracy": 0.13888985589146613,
      "num_tokens": 4263111.0,
      "step": 2455
    },
    {
      "entropy": 6.085035181045532,
      "epoch": 0.19140245088504182,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998546685907225,
      "loss": 5.9222,
      "mean_token_accuracy": 0.1431556224822998,
      "num_tokens": 4272796.0,
      "step": 2460
    },
    {
      "entropy": 6.06772747039795,
      "epoch": 0.19179148025675938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998536708891927,
      "loss": 5.853,
      "mean_token_accuracy": 0.1444825768470764,
      "num_tokens": 4281361.0,
      "step": 2465
    },
    {
      "entropy": 6.134342861175537,
      "epoch": 0.19218050962847694,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998526697758663,
      "loss": 5.909,
      "mean_token_accuracy": 0.13936054483056068,
      "num_tokens": 4290470.0,
      "step": 2470
    },
    {
      "entropy": 6.100904512405395,
      "epoch": 0.19256953900019452,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998516652507585,
      "loss": 5.9476,
      "mean_token_accuracy": 0.1369578666985035,
      "num_tokens": 4299087.0,
      "step": 2475
    },
    {
      "entropy": 6.05439133644104,
      "epoch": 0.19295856837191208,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998506573138846,
      "loss": 5.8166,
      "mean_token_accuracy": 0.1480974078178406,
      "num_tokens": 4307572.0,
      "step": 2480
    },
    {
      "entropy": 6.066846036911011,
      "epoch": 0.19334759774362964,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998496459652598,
      "loss": 5.9009,
      "mean_token_accuracy": 0.14306523129343987,
      "num_tokens": 4316337.0,
      "step": 2485
    },
    {
      "entropy": 6.15980863571167,
      "epoch": 0.1937366271153472,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998486312048995,
      "loss": 5.9834,
      "mean_token_accuracy": 0.14186982065439224,
      "num_tokens": 4326258.0,
      "step": 2490
    },
    {
      "entropy": 6.080019283294678,
      "epoch": 0.19412565648706478,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998476130328193,
      "loss": 5.842,
      "mean_token_accuracy": 0.14730110839009286,
      "num_tokens": 4334711.0,
      "step": 2495
    },
    {
      "entropy": 6.140625524520874,
      "epoch": 0.19451468585878234,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998465914490343,
      "loss": 5.9334,
      "mean_token_accuracy": 0.13907926082611083,
      "num_tokens": 4343739.0,
      "step": 2500
    },
    {
      "entropy": 6.078987503051758,
      "epoch": 0.1949037152304999,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049984556645356,
      "loss": 5.8735,
      "mean_token_accuracy": 0.14407971054315566,
      "num_tokens": 4352905.0,
      "step": 2505
    },
    {
      "entropy": 6.112984275817871,
      "epoch": 0.19529274460221746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998445380464124,
      "loss": 5.8965,
      "mean_token_accuracy": 0.14865672141313552,
      "num_tokens": 4361333.0,
      "step": 2510
    },
    {
      "entropy": 6.0111664772033695,
      "epoch": 0.19568177397393502,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998435062276066,
      "loss": 5.8578,
      "mean_token_accuracy": 0.14098201543092728,
      "num_tokens": 4369872.0,
      "step": 2515
    },
    {
      "entropy": 6.126198101043701,
      "epoch": 0.1960708033456526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998424709971586,
      "loss": 5.9367,
      "mean_token_accuracy": 0.14096542671322823,
      "num_tokens": 4378097.0,
      "step": 2520
    },
    {
      "entropy": 6.071844673156738,
      "epoch": 0.19645983271737016,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998414323550839,
      "loss": 5.856,
      "mean_token_accuracy": 0.14555131644010544,
      "num_tokens": 4387414.0,
      "step": 2525
    },
    {
      "entropy": 6.103514766693115,
      "epoch": 0.19684886208908772,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004998403903013984,
      "loss": 5.9399,
      "mean_token_accuracy": 0.14774652421474457,
      "num_tokens": 4395922.0,
      "step": 2530
    },
    {
      "entropy": 6.063558340072632,
      "epoch": 0.19723789146080528,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998393448361179,
      "loss": 5.9156,
      "mean_token_accuracy": 0.138917675614357,
      "num_tokens": 4404500.0,
      "step": 2535
    },
    {
      "entropy": 5.985238027572632,
      "epoch": 0.19762692083252287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499838295959258,
      "loss": 5.9028,
      "mean_token_accuracy": 0.1411334067583084,
      "num_tokens": 4413433.0,
      "step": 2540
    },
    {
      "entropy": 6.1477453231811525,
      "epoch": 0.19801595020424043,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499837243670835,
      "loss": 5.928,
      "mean_token_accuracy": 0.14450406208634375,
      "num_tokens": 4421789.0,
      "step": 2545
    },
    {
      "entropy": 6.044068193435669,
      "epoch": 0.19840497957595798,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998361879708646,
      "loss": 5.7949,
      "mean_token_accuracy": 0.15504934936761855,
      "num_tokens": 4430256.0,
      "step": 2550
    },
    {
      "entropy": 6.045944166183472,
      "epoch": 0.19879400894767554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499835128859363,
      "loss": 5.8457,
      "mean_token_accuracy": 0.1446925073862076,
      "num_tokens": 4438431.0,
      "step": 2555
    },
    {
      "entropy": 6.091934299468994,
      "epoch": 0.1991830383193931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998340663363461,
      "loss": 5.9122,
      "mean_token_accuracy": 0.1388619139790535,
      "num_tokens": 4447150.0,
      "step": 2560
    },
    {
      "entropy": 6.038812255859375,
      "epoch": 0.1995720676911107,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998330004018301,
      "loss": 5.7921,
      "mean_token_accuracy": 0.14743948429822923,
      "num_tokens": 4455181.0,
      "step": 2565
    },
    {
      "entropy": 6.0153632164001465,
      "epoch": 0.19996109706282825,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998319310558312,
      "loss": 5.8176,
      "mean_token_accuracy": 0.1437955118715763,
      "num_tokens": 4463361.0,
      "step": 2570
    },
    {
      "entropy": 6.06419997215271,
      "epoch": 0.2003501264345458,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998308582983657,
      "loss": 5.8634,
      "mean_token_accuracy": 0.14697861596941947,
      "num_tokens": 4471853.0,
      "step": 2575
    },
    {
      "entropy": 5.983008098602295,
      "epoch": 0.20073915580626336,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998297821294497,
      "loss": 5.8164,
      "mean_token_accuracy": 0.14654360562562943,
      "num_tokens": 4480812.0,
      "step": 2580
    },
    {
      "entropy": 6.073977756500244,
      "epoch": 0.20112818517798095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998287025490995,
      "loss": 5.8683,
      "mean_token_accuracy": 0.1451071783900261,
      "num_tokens": 4489180.0,
      "step": 2585
    },
    {
      "entropy": 6.090311336517334,
      "epoch": 0.2015172145496985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998276195573317,
      "loss": 5.9298,
      "mean_token_accuracy": 0.14020900428295135,
      "num_tokens": 4498013.0,
      "step": 2590
    },
    {
      "entropy": 6.161534404754638,
      "epoch": 0.20190624392141607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998265331541627,
      "loss": 5.9647,
      "mean_token_accuracy": 0.1288097895681858,
      "num_tokens": 4506767.0,
      "step": 2595
    },
    {
      "entropy": 6.1162800788879395,
      "epoch": 0.20229527329313363,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998254433396088,
      "loss": 5.9698,
      "mean_token_accuracy": 0.14163854718208313,
      "num_tokens": 4516657.0,
      "step": 2600
    },
    {
      "entropy": 5.995081090927124,
      "epoch": 0.20268430266485118,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998243501136867,
      "loss": 5.8093,
      "mean_token_accuracy": 0.14900346398353576,
      "num_tokens": 4524980.0,
      "step": 2605
    },
    {
      "entropy": 6.035889768600464,
      "epoch": 0.20307333203656877,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998232534764129,
      "loss": 5.8681,
      "mean_token_accuracy": 0.14729219824075698,
      "num_tokens": 4533236.0,
      "step": 2610
    },
    {
      "entropy": 5.970295190811157,
      "epoch": 0.20346236140828633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998221534278041,
      "loss": 5.7612,
      "mean_token_accuracy": 0.1461296297609806,
      "num_tokens": 4541778.0,
      "step": 2615
    },
    {
      "entropy": 6.136829423904419,
      "epoch": 0.2038513907800039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998210499678769,
      "loss": 5.8873,
      "mean_token_accuracy": 0.14116985723376274,
      "num_tokens": 4551131.0,
      "step": 2620
    },
    {
      "entropy": 6.106343078613281,
      "epoch": 0.20424042015172145,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998199430966481,
      "loss": 5.8658,
      "mean_token_accuracy": 0.1418320581316948,
      "num_tokens": 4559530.0,
      "step": 2625
    },
    {
      "entropy": 5.890627050399781,
      "epoch": 0.20462944952343903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998188328141346,
      "loss": 5.7583,
      "mean_token_accuracy": 0.14722466468811035,
      "num_tokens": 4567533.0,
      "step": 2630
    },
    {
      "entropy": 6.181466436386108,
      "epoch": 0.2050184788951566,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998177191203531,
      "loss": 5.896,
      "mean_token_accuracy": 0.13903536647558212,
      "num_tokens": 4576414.0,
      "step": 2635
    },
    {
      "entropy": 6.053440952301026,
      "epoch": 0.20540750826687415,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998166020153204,
      "loss": 5.8382,
      "mean_token_accuracy": 0.147342050075531,
      "num_tokens": 4585331.0,
      "step": 2640
    },
    {
      "entropy": 6.052559661865234,
      "epoch": 0.2057965376385917,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998154814990538,
      "loss": 5.8647,
      "mean_token_accuracy": 0.14799018651247026,
      "num_tokens": 4593526.0,
      "step": 2645
    },
    {
      "entropy": 6.026284027099609,
      "epoch": 0.20618556701030927,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049981435757157,
      "loss": 5.8634,
      "mean_token_accuracy": 0.14488066285848616,
      "num_tokens": 4602400.0,
      "step": 2650
    },
    {
      "entropy": 5.989109945297241,
      "epoch": 0.20657459638202685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998132302328862,
      "loss": 5.8003,
      "mean_token_accuracy": 0.14428150430321693,
      "num_tokens": 4611037.0,
      "step": 2655
    },
    {
      "entropy": 6.0565345764160154,
      "epoch": 0.2069636257537444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998120994830195,
      "loss": 5.8115,
      "mean_token_accuracy": 0.14851100668311118,
      "num_tokens": 4619030.0,
      "step": 2660
    },
    {
      "entropy": 6.032034778594971,
      "epoch": 0.20735265512546197,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499810965321987,
      "loss": 5.8228,
      "mean_token_accuracy": 0.14074279814958573,
      "num_tokens": 4628649.0,
      "step": 2665
    },
    {
      "entropy": 5.993114280700683,
      "epoch": 0.20774168449717953,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998098277498059,
      "loss": 5.7309,
      "mean_token_accuracy": 0.14714966639876365,
      "num_tokens": 4636935.0,
      "step": 2670
    },
    {
      "entropy": 6.017863130569458,
      "epoch": 0.2081307138688971,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998086867664934,
      "loss": 5.9197,
      "mean_token_accuracy": 0.14219892248511315,
      "num_tokens": 4645532.0,
      "step": 2675
    },
    {
      "entropy": 6.1221582889556885,
      "epoch": 0.20851974324061467,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499807542372067,
      "loss": 5.9851,
      "mean_token_accuracy": 0.13557760491967202,
      "num_tokens": 4654433.0,
      "step": 2680
    },
    {
      "entropy": 6.139663934707642,
      "epoch": 0.20890877261233223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998063945665439,
      "loss": 5.989,
      "mean_token_accuracy": 0.13590591549873351,
      "num_tokens": 4662913.0,
      "step": 2685
    },
    {
      "entropy": 6.098662042617798,
      "epoch": 0.2092978019840498,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998052433499416,
      "loss": 5.9612,
      "mean_token_accuracy": 0.1359117828309536,
      "num_tokens": 4672270.0,
      "step": 2690
    },
    {
      "entropy": 6.16609034538269,
      "epoch": 0.20968683135576735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998040887222776,
      "loss": 5.9854,
      "mean_token_accuracy": 0.1398947387933731,
      "num_tokens": 4681428.0,
      "step": 2695
    },
    {
      "entropy": 6.0872087478637695,
      "epoch": 0.21007586072748494,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998029306835693,
      "loss": 5.8204,
      "mean_token_accuracy": 0.14755907729268075,
      "num_tokens": 4690492.0,
      "step": 2700
    },
    {
      "entropy": 6.07254843711853,
      "epoch": 0.2104648900992025,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998017692338344,
      "loss": 5.9404,
      "mean_token_accuracy": 0.14041918218135835,
      "num_tokens": 4699667.0,
      "step": 2705
    },
    {
      "entropy": 5.926076602935791,
      "epoch": 0.21085391947092005,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998006043730905,
      "loss": 5.6251,
      "mean_token_accuracy": 0.1608228325843811,
      "num_tokens": 4708529.0,
      "step": 2710
    },
    {
      "entropy": 5.902212381362915,
      "epoch": 0.2112429488426376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997994361013551,
      "loss": 5.8837,
      "mean_token_accuracy": 0.13886216133832932,
      "num_tokens": 4717684.0,
      "step": 2715
    },
    {
      "entropy": 6.049212312698364,
      "epoch": 0.21163197821435517,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997982644186461,
      "loss": 5.8008,
      "mean_token_accuracy": 0.14885463267564775,
      "num_tokens": 4727067.0,
      "step": 2720
    },
    {
      "entropy": 6.025175905227661,
      "epoch": 0.21202100758607276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997970893249813,
      "loss": 5.7868,
      "mean_token_accuracy": 0.14974473118782045,
      "num_tokens": 4735434.0,
      "step": 2725
    },
    {
      "entropy": 6.024233770370484,
      "epoch": 0.21241003695779032,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997959108203785,
      "loss": 5.8227,
      "mean_token_accuracy": 0.14492656514048577,
      "num_tokens": 4743843.0,
      "step": 2730
    },
    {
      "entropy": 6.081152391433716,
      "epoch": 0.21279906632950787,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997947289048554,
      "loss": 5.8634,
      "mean_token_accuracy": 0.1460092097520828,
      "num_tokens": 4753228.0,
      "step": 2735
    },
    {
      "entropy": 6.036781454086304,
      "epoch": 0.21318809570122543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997935435784302,
      "loss": 5.9396,
      "mean_token_accuracy": 0.1427992545068264,
      "num_tokens": 4763401.0,
      "step": 2740
    },
    {
      "entropy": 6.008918571472168,
      "epoch": 0.21357712507294302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997923548411208,
      "loss": 5.8139,
      "mean_token_accuracy": 0.1405964434146881,
      "num_tokens": 4771812.0,
      "step": 2745
    },
    {
      "entropy": 5.9517346858978275,
      "epoch": 0.21396615444466058,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499791162692945,
      "loss": 5.835,
      "mean_token_accuracy": 0.14680606722831727,
      "num_tokens": 4780616.0,
      "step": 2750
    },
    {
      "entropy": 5.967705154418946,
      "epoch": 0.21435518381637814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997899671339214,
      "loss": 5.8298,
      "mean_token_accuracy": 0.1467273563146591,
      "num_tokens": 4789886.0,
      "step": 2755
    },
    {
      "entropy": 6.010776853561401,
      "epoch": 0.2147442131880957,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997887681640675,
      "loss": 5.7924,
      "mean_token_accuracy": 0.15051775872707368,
      "num_tokens": 4799189.0,
      "step": 2760
    },
    {
      "entropy": 6.061824893951416,
      "epoch": 0.21513324255981325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997875657834021,
      "loss": 5.8594,
      "mean_token_accuracy": 0.14805168360471727,
      "num_tokens": 4807850.0,
      "step": 2765
    },
    {
      "entropy": 6.05089316368103,
      "epoch": 0.21552227193153084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997863599919432,
      "loss": 5.7707,
      "mean_token_accuracy": 0.14720831513404847,
      "num_tokens": 4815742.0,
      "step": 2770
    },
    {
      "entropy": 6.032073307037353,
      "epoch": 0.2159113013032484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997851507897089,
      "loss": 5.857,
      "mean_token_accuracy": 0.14828907027840615,
      "num_tokens": 4824761.0,
      "step": 2775
    },
    {
      "entropy": 6.07795557975769,
      "epoch": 0.21630033067496596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997839381767178,
      "loss": 5.9915,
      "mean_token_accuracy": 0.1347334012389183,
      "num_tokens": 4832805.0,
      "step": 2780
    },
    {
      "entropy": 5.988602924346924,
      "epoch": 0.21668936004668352,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997827221529882,
      "loss": 5.8384,
      "mean_token_accuracy": 0.1470751255750656,
      "num_tokens": 4841938.0,
      "step": 2785
    },
    {
      "entropy": 6.045064449310303,
      "epoch": 0.2170783894184011,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997815027185386,
      "loss": 5.8645,
      "mean_token_accuracy": 0.1436170071363449,
      "num_tokens": 4849932.0,
      "step": 2790
    },
    {
      "entropy": 5.966535568237305,
      "epoch": 0.21746741879011866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997802798733874,
      "loss": 5.7285,
      "mean_token_accuracy": 0.1491599291563034,
      "num_tokens": 4859054.0,
      "step": 2795
    },
    {
      "entropy": 6.030629825592041,
      "epoch": 0.21785644816183622,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997790536175534,
      "loss": 5.8931,
      "mean_token_accuracy": 0.15091775506734847,
      "num_tokens": 4867415.0,
      "step": 2800
    },
    {
      "entropy": 6.073462200164795,
      "epoch": 0.21824547753355378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997778239510548,
      "loss": 5.9334,
      "mean_token_accuracy": 0.14272018224000932,
      "num_tokens": 4875965.0,
      "step": 2805
    },
    {
      "entropy": 6.065163946151733,
      "epoch": 0.21863450690527134,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997765908739106,
      "loss": 5.8328,
      "mean_token_accuracy": 0.14725763201713563,
      "num_tokens": 4884629.0,
      "step": 2810
    },
    {
      "entropy": 6.022565269470215,
      "epoch": 0.21902353627698892,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997753543861395,
      "loss": 5.8746,
      "mean_token_accuracy": 0.14686051681637763,
      "num_tokens": 4893573.0,
      "step": 2815
    },
    {
      "entropy": 5.940307140350342,
      "epoch": 0.21941256564870648,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049977411448776,
      "loss": 5.8078,
      "mean_token_accuracy": 0.1442374676465988,
      "num_tokens": 4902831.0,
      "step": 2820
    },
    {
      "entropy": 6.07591495513916,
      "epoch": 0.21980159502042404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997728711787912,
      "loss": 5.8593,
      "mean_token_accuracy": 0.13736442774534224,
      "num_tokens": 4911360.0,
      "step": 2825
    },
    {
      "entropy": 6.043661260604859,
      "epoch": 0.2201906243921416,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997716244592518,
      "loss": 5.8214,
      "mean_token_accuracy": 0.14348886758089066,
      "num_tokens": 4919758.0,
      "step": 2830
    },
    {
      "entropy": 5.896661329269409,
      "epoch": 0.22057965376385918,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997703743291607,
      "loss": 5.7102,
      "mean_token_accuracy": 0.15584124326705934,
      "num_tokens": 4928934.0,
      "step": 2835
    },
    {
      "entropy": 5.933321142196656,
      "epoch": 0.22096868313557674,
      "grad_norm": 1.0,
      "learning_rate": 0.000499769120788537,
      "loss": 5.7655,
      "mean_token_accuracy": 0.14841632395982743,
      "num_tokens": 4937780.0,
      "step": 2840
    },
    {
      "entropy": 6.14625506401062,
      "epoch": 0.2213577125072943,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997678638373995,
      "loss": 5.8583,
      "mean_token_accuracy": 0.1424436628818512,
      "num_tokens": 4946862.0,
      "step": 2845
    },
    {
      "entropy": 5.994505119323731,
      "epoch": 0.22174674187901186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997666034757676,
      "loss": 5.8358,
      "mean_token_accuracy": 0.149576772749424,
      "num_tokens": 4954883.0,
      "step": 2850
    },
    {
      "entropy": 5.885545825958252,
      "epoch": 0.22213577125072942,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997653397036603,
      "loss": 5.6416,
      "mean_token_accuracy": 0.1590978369116783,
      "num_tokens": 4962840.0,
      "step": 2855
    },
    {
      "entropy": 5.9540290355682375,
      "epoch": 0.222524800622447,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997640725210965,
      "loss": 5.8776,
      "mean_token_accuracy": 0.14788952842354774,
      "num_tokens": 4972334.0,
      "step": 2860
    },
    {
      "entropy": 6.020223808288574,
      "epoch": 0.22291382999416456,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997628019280958,
      "loss": 5.8327,
      "mean_token_accuracy": 0.14812421351671218,
      "num_tokens": 4981032.0,
      "step": 2865
    },
    {
      "entropy": 6.030521392822266,
      "epoch": 0.22330285936588212,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997615279246773,
      "loss": 5.8324,
      "mean_token_accuracy": 0.14387644529342652,
      "num_tokens": 4990502.0,
      "step": 2870
    },
    {
      "entropy": 6.103343868255616,
      "epoch": 0.22369188873759968,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997602505108603,
      "loss": 6.0219,
      "mean_token_accuracy": 0.1379172571003437,
      "num_tokens": 4999121.0,
      "step": 2875
    },
    {
      "entropy": 6.037298774719238,
      "epoch": 0.22408091810931724,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997589696866644,
      "loss": 5.7702,
      "mean_token_accuracy": 0.14713070988655091,
      "num_tokens": 5007271.0,
      "step": 2880
    },
    {
      "entropy": 5.960604190826416,
      "epoch": 0.22446994748103483,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997576854521088,
      "loss": 5.8223,
      "mean_token_accuracy": 0.1438037596642971,
      "num_tokens": 5015940.0,
      "step": 2885
    },
    {
      "entropy": 5.943349742889405,
      "epoch": 0.22485897685275238,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499756397807213,
      "loss": 5.7488,
      "mean_token_accuracy": 0.15114528685808182,
      "num_tokens": 5024294.0,
      "step": 2890
    },
    {
      "entropy": 6.015953445434571,
      "epoch": 0.22524800622446994,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997551067519966,
      "loss": 5.7094,
      "mean_token_accuracy": 0.1564256578683853,
      "num_tokens": 5033401.0,
      "step": 2895
    },
    {
      "entropy": 5.932667875289917,
      "epoch": 0.2256370355961875,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997538122864792,
      "loss": 5.8063,
      "mean_token_accuracy": 0.14516399949789047,
      "num_tokens": 5041955.0,
      "step": 2900
    },
    {
      "entropy": 5.982646894454956,
      "epoch": 0.2260260649679051,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997525144106804,
      "loss": 5.8437,
      "mean_token_accuracy": 0.14041123390197754,
      "num_tokens": 5051132.0,
      "step": 2905
    },
    {
      "entropy": 5.9859472751617435,
      "epoch": 0.22641509433962265,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049975121312462,
      "loss": 5.7668,
      "mean_token_accuracy": 0.15441568344831466,
      "num_tokens": 5059492.0,
      "step": 2910
    },
    {
      "entropy": 5.891498994827271,
      "epoch": 0.2268041237113402,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997499084283177,
      "loss": 5.8123,
      "mean_token_accuracy": 0.15069894939661027,
      "num_tokens": 5069015.0,
      "step": 2915
    },
    {
      "entropy": 5.9988916397094725,
      "epoch": 0.22719315308305776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997486003217932,
      "loss": 5.8223,
      "mean_token_accuracy": 0.1472141906619072,
      "num_tokens": 5077614.0,
      "step": 2920
    },
    {
      "entropy": 5.993191337585449,
      "epoch": 0.22758218245477532,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997472888050664,
      "loss": 5.7616,
      "mean_token_accuracy": 0.14898982793092727,
      "num_tokens": 5086300.0,
      "step": 2925
    },
    {
      "entropy": 5.879473543167114,
      "epoch": 0.2279712118264929,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997459738781573,
      "loss": 5.8034,
      "mean_token_accuracy": 0.1515287682414055,
      "num_tokens": 5095967.0,
      "step": 2930
    },
    {
      "entropy": 6.005168533325195,
      "epoch": 0.22836024119821047,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997446555410857,
      "loss": 5.7629,
      "mean_token_accuracy": 0.14633936136960984,
      "num_tokens": 5104611.0,
      "step": 2935
    },
    {
      "entropy": 5.9345582008361815,
      "epoch": 0.22874927056992803,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997433337938716,
      "loss": 5.7352,
      "mean_token_accuracy": 0.14881281405687333,
      "num_tokens": 5113452.0,
      "step": 2940
    },
    {
      "entropy": 5.903948497772217,
      "epoch": 0.22913829994164558,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997420086365351,
      "loss": 5.7075,
      "mean_token_accuracy": 0.15194878876209258,
      "num_tokens": 5121558.0,
      "step": 2945
    },
    {
      "entropy": 5.9376458644866945,
      "epoch": 0.22952732931336317,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997406800690966,
      "loss": 5.6743,
      "mean_token_accuracy": 0.16064000129699707,
      "num_tokens": 5129942.0,
      "step": 2950
    },
    {
      "entropy": 5.969931030273438,
      "epoch": 0.22991635868508073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997393480915758,
      "loss": 5.8154,
      "mean_token_accuracy": 0.14137779250741006,
      "num_tokens": 5139012.0,
      "step": 2955
    },
    {
      "entropy": 6.0334978103637695,
      "epoch": 0.2303053880567983,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997380127039931,
      "loss": 5.818,
      "mean_token_accuracy": 0.150064118206501,
      "num_tokens": 5148185.0,
      "step": 2960
    },
    {
      "entropy": 5.993949317932129,
      "epoch": 0.23069441742851585,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997366739063688,
      "loss": 5.8355,
      "mean_token_accuracy": 0.1448197141289711,
      "num_tokens": 5157086.0,
      "step": 2965
    },
    {
      "entropy": 6.100489139556885,
      "epoch": 0.2310834468002334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997353316987231,
      "loss": 5.9076,
      "mean_token_accuracy": 0.14102671518921853,
      "num_tokens": 5165842.0,
      "step": 2970
    },
    {
      "entropy": 6.074266862869263,
      "epoch": 0.231472476171951,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997339860810765,
      "loss": 5.9076,
      "mean_token_accuracy": 0.14826976209878923,
      "num_tokens": 5174161.0,
      "step": 2975
    },
    {
      "entropy": 5.951193332672119,
      "epoch": 0.23186150554366855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997326370534495,
      "loss": 5.8456,
      "mean_token_accuracy": 0.14642503708600998,
      "num_tokens": 5183830.0,
      "step": 2980
    },
    {
      "entropy": 5.981384754180908,
      "epoch": 0.2322505349153861,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997312846158622,
      "loss": 5.7689,
      "mean_token_accuracy": 0.14501096159219742,
      "num_tokens": 5192248.0,
      "step": 2985
    },
    {
      "entropy": 5.878167724609375,
      "epoch": 0.23263956428710367,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997299287683355,
      "loss": 5.7312,
      "mean_token_accuracy": 0.1538401484489441,
      "num_tokens": 5201212.0,
      "step": 2990
    },
    {
      "entropy": 5.966686391830445,
      "epoch": 0.23302859365882125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997285695108898,
      "loss": 5.7912,
      "mean_token_accuracy": 0.145901557803154,
      "num_tokens": 5209338.0,
      "step": 2995
    },
    {
      "entropy": 6.075098180770874,
      "epoch": 0.2334176230305388,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997272068435458,
      "loss": 5.8133,
      "mean_token_accuracy": 0.14372822940349578,
      "num_tokens": 5217590.0,
      "step": 3000
    },
    {
      "epoch": 0.2334176230305388,
      "eval_entropy": 5.7649864601561145,
      "eval_loss": 5.838047027587891,
      "eval_mean_token_accuracy": 0.1541939339104814,
      "eval_num_tokens": 5217590.0,
      "eval_runtime": 21.9076,
      "eval_samples_per_second": 1896.968,
      "eval_steps_per_second": 237.132,
      "step": 3000
    },
    {
      "entropy": 5.907944631576538,
      "epoch": 0.23380665240225637,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499725840766324,
      "loss": 5.7253,
      "mean_token_accuracy": 0.15077031999826432,
      "num_tokens": 5226383.0,
      "step": 3005
    },
    {
      "entropy": 5.9791473865509035,
      "epoch": 0.23419568177397393,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997244712792453,
      "loss": 5.7687,
      "mean_token_accuracy": 0.1488740026950836,
      "num_tokens": 5235517.0,
      "step": 3010
    },
    {
      "entropy": 6.0556151390075685,
      "epoch": 0.2345847111456915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997230983823305,
      "loss": 5.7481,
      "mean_token_accuracy": 0.1573214754462242,
      "num_tokens": 5243812.0,
      "step": 3015
    },
    {
      "entropy": 5.912490129470825,
      "epoch": 0.23497374051740907,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997217220756003,
      "loss": 5.7598,
      "mean_token_accuracy": 0.14918334409594536,
      "num_tokens": 5252399.0,
      "step": 3020
    },
    {
      "entropy": 5.963000392913818,
      "epoch": 0.23536276988912663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997203423590757,
      "loss": 5.7813,
      "mean_token_accuracy": 0.14887740164995195,
      "num_tokens": 5260847.0,
      "step": 3025
    },
    {
      "entropy": 5.998490476608277,
      "epoch": 0.2357517992608442,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997189592327775,
      "loss": 5.855,
      "mean_token_accuracy": 0.14860740453004836,
      "num_tokens": 5268847.0,
      "step": 3030
    },
    {
      "entropy": 6.0372459411621096,
      "epoch": 0.23614082863256175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997175726967268,
      "loss": 5.7837,
      "mean_token_accuracy": 0.14813946485519408,
      "num_tokens": 5277989.0,
      "step": 3035
    },
    {
      "entropy": 5.881270456314087,
      "epoch": 0.23652985800427934,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997161827509447,
      "loss": 5.7804,
      "mean_token_accuracy": 0.15062111616134644,
      "num_tokens": 5286272.0,
      "step": 3040
    },
    {
      "entropy": 5.935083866119385,
      "epoch": 0.2369188873759969,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997147893954521,
      "loss": 5.7856,
      "mean_token_accuracy": 0.14935739189386368,
      "num_tokens": 5295281.0,
      "step": 3045
    },
    {
      "entropy": 6.040230655670166,
      "epoch": 0.23730791674771445,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997133926302702,
      "loss": 5.7326,
      "mean_token_accuracy": 0.15305247008800507,
      "num_tokens": 5303905.0,
      "step": 3050
    },
    {
      "entropy": 5.91657977104187,
      "epoch": 0.237696946119432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997119924554204,
      "loss": 5.7193,
      "mean_token_accuracy": 0.1474948540329933,
      "num_tokens": 5313146.0,
      "step": 3055
    },
    {
      "entropy": 5.920493745803833,
      "epoch": 0.23808597549114957,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997105888709236,
      "loss": 5.755,
      "mean_token_accuracy": 0.14988460689783095,
      "num_tokens": 5320838.0,
      "step": 3060
    },
    {
      "entropy": 5.927872037887573,
      "epoch": 0.23847500486286716,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997091818768015,
      "loss": 5.7825,
      "mean_token_accuracy": 0.14742809981107713,
      "num_tokens": 5329312.0,
      "step": 3065
    },
    {
      "entropy": 5.909237957000732,
      "epoch": 0.23886403423458472,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499707771473075,
      "loss": 5.777,
      "mean_token_accuracy": 0.14921307563781738,
      "num_tokens": 5337747.0,
      "step": 3070
    },
    {
      "entropy": 5.979187774658203,
      "epoch": 0.23925306360630227,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997063576597659,
      "loss": 5.7621,
      "mean_token_accuracy": 0.1467119038105011,
      "num_tokens": 5346928.0,
      "step": 3075
    },
    {
      "entropy": 5.943646335601807,
      "epoch": 0.23964209297801983,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997049404368954,
      "loss": 5.7875,
      "mean_token_accuracy": 0.1435694396495819,
      "num_tokens": 5355827.0,
      "step": 3080
    },
    {
      "entropy": 6.062672805786133,
      "epoch": 0.2400311223497374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499703519804485,
      "loss": 5.9124,
      "mean_token_accuracy": 0.14042111486196518,
      "num_tokens": 5364309.0,
      "step": 3085
    },
    {
      "entropy": 5.996913576126099,
      "epoch": 0.24042015172145498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997020957625564,
      "loss": 5.8024,
      "mean_token_accuracy": 0.14950317591428758,
      "num_tokens": 5372979.0,
      "step": 3090
    },
    {
      "entropy": 5.954580926895142,
      "epoch": 0.24080918109317254,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997006683111312,
      "loss": 5.7816,
      "mean_token_accuracy": 0.1460615485906601,
      "num_tokens": 5381809.0,
      "step": 3095
    },
    {
      "entropy": 6.0027283191680905,
      "epoch": 0.2411982104648901,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499699237450231,
      "loss": 5.8266,
      "mean_token_accuracy": 0.1532255783677101,
      "num_tokens": 5390036.0,
      "step": 3100
    },
    {
      "entropy": 6.103567266464234,
      "epoch": 0.24158723983660765,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996978031798774,
      "loss": 5.8904,
      "mean_token_accuracy": 0.14301448315382004,
      "num_tokens": 5398739.0,
      "step": 3105
    },
    {
      "entropy": 5.922081136703492,
      "epoch": 0.24197626920832524,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996963655000924,
      "loss": 5.8034,
      "mean_token_accuracy": 0.14707928150892258,
      "num_tokens": 5407441.0,
      "step": 3110
    },
    {
      "entropy": 5.998806762695312,
      "epoch": 0.2423652985800428,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996949244108977,
      "loss": 5.7393,
      "mean_token_accuracy": 0.15036838203668595,
      "num_tokens": 5416534.0,
      "step": 3115
    },
    {
      "entropy": 5.997593593597412,
      "epoch": 0.24275432795176036,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499693479912315,
      "loss": 5.7991,
      "mean_token_accuracy": 0.14814602732658386,
      "num_tokens": 5425170.0,
      "step": 3120
    },
    {
      "entropy": 5.911679410934449,
      "epoch": 0.24314335732347792,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996920320043666,
      "loss": 5.7656,
      "mean_token_accuracy": 0.1437433570623398,
      "num_tokens": 5433721.0,
      "step": 3125
    },
    {
      "entropy": 6.002860069274902,
      "epoch": 0.24353238669519547,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996905806870742,
      "loss": 5.8225,
      "mean_token_accuracy": 0.14360172301530838,
      "num_tokens": 5441708.0,
      "step": 3130
    },
    {
      "entropy": 5.928072738647461,
      "epoch": 0.24392141606691306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996891259604598,
      "loss": 5.7297,
      "mean_token_accuracy": 0.15173338651657103,
      "num_tokens": 5449942.0,
      "step": 3135
    },
    {
      "entropy": 5.993139457702637,
      "epoch": 0.24431044543863062,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996876678245455,
      "loss": 5.751,
      "mean_token_accuracy": 0.14968564808368684,
      "num_tokens": 5458004.0,
      "step": 3140
    },
    {
      "entropy": 6.0081267833709715,
      "epoch": 0.24469947481034818,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996862062793536,
      "loss": 5.8396,
      "mean_token_accuracy": 0.15170659869909286,
      "num_tokens": 5467071.0,
      "step": 3145
    },
    {
      "entropy": 5.93834810256958,
      "epoch": 0.24508850418206574,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996847413249061,
      "loss": 5.8085,
      "mean_token_accuracy": 0.14909462928771972,
      "num_tokens": 5476221.0,
      "step": 3150
    },
    {
      "entropy": 5.976137924194336,
      "epoch": 0.24547753355378332,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996832729612252,
      "loss": 5.7315,
      "mean_token_accuracy": 0.14793778508901595,
      "num_tokens": 5484549.0,
      "step": 3155
    },
    {
      "entropy": 5.853325271606446,
      "epoch": 0.24586656292550088,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996818011883333,
      "loss": 5.7291,
      "mean_token_accuracy": 0.15456597805023192,
      "num_tokens": 5492909.0,
      "step": 3160
    },
    {
      "entropy": 5.959857606887818,
      "epoch": 0.24625559229721844,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996803260062527,
      "loss": 5.7712,
      "mean_token_accuracy": 0.15276994407176972,
      "num_tokens": 5501418.0,
      "step": 3165
    },
    {
      "entropy": 5.982174873352051,
      "epoch": 0.246644621668936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996788474150057,
      "loss": 5.7809,
      "mean_token_accuracy": 0.14409117698669432,
      "num_tokens": 5510609.0,
      "step": 3170
    },
    {
      "entropy": 5.861660671234131,
      "epoch": 0.24703365104065356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996773654146149,
      "loss": 5.6293,
      "mean_token_accuracy": 0.15756507366895675,
      "num_tokens": 5518349.0,
      "step": 3175
    },
    {
      "entropy": 5.857954120635986,
      "epoch": 0.24742268041237114,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996758800051026,
      "loss": 5.5799,
      "mean_token_accuracy": 0.15771122723817826,
      "num_tokens": 5526972.0,
      "step": 3180
    },
    {
      "entropy": 5.867163467407226,
      "epoch": 0.2478117097840887,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996743911864916,
      "loss": 5.752,
      "mean_token_accuracy": 0.14744784981012343,
      "num_tokens": 5535683.0,
      "step": 3185
    },
    {
      "entropy": 5.870160245895386,
      "epoch": 0.24820073915580626,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996728989588041,
      "loss": 5.7191,
      "mean_token_accuracy": 0.1577271744608879,
      "num_tokens": 5544258.0,
      "step": 3190
    },
    {
      "entropy": 5.846530818939209,
      "epoch": 0.24858976852752382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996714033220632,
      "loss": 5.7347,
      "mean_token_accuracy": 0.14929380863904954,
      "num_tokens": 5552608.0,
      "step": 3195
    },
    {
      "entropy": 5.937586975097656,
      "epoch": 0.2489787978992414,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996699042762911,
      "loss": 5.8503,
      "mean_token_accuracy": 0.14045757427811623,
      "num_tokens": 5561357.0,
      "step": 3200
    },
    {
      "entropy": 6.005559206008911,
      "epoch": 0.24936782727095896,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996684018215111,
      "loss": 5.6715,
      "mean_token_accuracy": 0.16094748824834823,
      "num_tokens": 5569862.0,
      "step": 3205
    },
    {
      "entropy": 5.907440137863159,
      "epoch": 0.24975685664267652,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996668959577453,
      "loss": 5.7432,
      "mean_token_accuracy": 0.15263464897871018,
      "num_tokens": 5578485.0,
      "step": 3210
    },
    {
      "entropy": 5.875484657287598,
      "epoch": 0.2501458860143941,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996653866850173,
      "loss": 5.7328,
      "mean_token_accuracy": 0.14682287722826004,
      "num_tokens": 5587713.0,
      "step": 3215
    },
    {
      "entropy": 5.963264131546021,
      "epoch": 0.25053491538611167,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996638740033495,
      "loss": 5.7997,
      "mean_token_accuracy": 0.15026128888130189,
      "num_tokens": 5596541.0,
      "step": 3220
    },
    {
      "entropy": 5.863694381713867,
      "epoch": 0.2509239447578292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996623579127651,
      "loss": 5.6721,
      "mean_token_accuracy": 0.15386685878038406,
      "num_tokens": 5605573.0,
      "step": 3225
    },
    {
      "entropy": 5.923236894607544,
      "epoch": 0.2513129741295468,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996608384132868,
      "loss": 5.6887,
      "mean_token_accuracy": 0.15542172491550446,
      "num_tokens": 5614016.0,
      "step": 3230
    },
    {
      "entropy": 6.036375427246094,
      "epoch": 0.25170200350126437,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499659315504938,
      "loss": 5.9137,
      "mean_token_accuracy": 0.1452291265130043,
      "num_tokens": 5622451.0,
      "step": 3235
    },
    {
      "entropy": 5.9037364482879635,
      "epoch": 0.2520910328729819,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004996577891877417,
      "loss": 5.7515,
      "mean_token_accuracy": 0.14964452758431435,
      "num_tokens": 5631384.0,
      "step": 3240
    },
    {
      "entropy": 5.927656078338623,
      "epoch": 0.2524800622446995,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996562594617209,
      "loss": 5.7129,
      "mean_token_accuracy": 0.1541423261165619,
      "num_tokens": 5640726.0,
      "step": 3245
    },
    {
      "entropy": 5.965122318267822,
      "epoch": 0.252869091616417,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996547263268991,
      "loss": 5.7332,
      "mean_token_accuracy": 0.14803583323955535,
      "num_tokens": 5648948.0,
      "step": 3250
    },
    {
      "entropy": 5.794720411300659,
      "epoch": 0.2532581209881346,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996531897832993,
      "loss": 5.6824,
      "mean_token_accuracy": 0.15577371567487716,
      "num_tokens": 5657951.0,
      "step": 3255
    },
    {
      "entropy": 6.0227070331573485,
      "epoch": 0.2536471503598522,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499651649830945,
      "loss": 5.8137,
      "mean_token_accuracy": 0.1499185547232628,
      "num_tokens": 5667476.0,
      "step": 3260
    },
    {
      "entropy": 5.9482104778289795,
      "epoch": 0.2540361797315697,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996501064698593,
      "loss": 5.8079,
      "mean_token_accuracy": 0.1510738968849182,
      "num_tokens": 5676627.0,
      "step": 3265
    },
    {
      "entropy": 5.888298082351684,
      "epoch": 0.2544252091032873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499648559700066,
      "loss": 5.7389,
      "mean_token_accuracy": 0.15606578439474106,
      "num_tokens": 5684376.0,
      "step": 3270
    },
    {
      "entropy": 5.999397611618042,
      "epoch": 0.25481423847500484,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996470095215884,
      "loss": 5.8974,
      "mean_token_accuracy": 0.13699436113238334,
      "num_tokens": 5693629.0,
      "step": 3275
    },
    {
      "entropy": 6.0265604019165036,
      "epoch": 0.2552032678467224,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996454559344498,
      "loss": 5.7848,
      "mean_token_accuracy": 0.1482024610042572,
      "num_tokens": 5702010.0,
      "step": 3280
    },
    {
      "entropy": 5.836571741104126,
      "epoch": 0.25559229721844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996438989386741,
      "loss": 5.6821,
      "mean_token_accuracy": 0.15332212150096894,
      "num_tokens": 5710823.0,
      "step": 3285
    },
    {
      "entropy": 5.929405069351196,
      "epoch": 0.25598132659015754,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996423385342847,
      "loss": 5.7385,
      "mean_token_accuracy": 0.15431096404790878,
      "num_tokens": 5719389.0,
      "step": 3290
    },
    {
      "entropy": 5.9652153015136715,
      "epoch": 0.25637035596187513,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996407747213055,
      "loss": 5.7513,
      "mean_token_accuracy": 0.14783500358462334,
      "num_tokens": 5727327.0,
      "step": 3295
    },
    {
      "entropy": 5.902391672134399,
      "epoch": 0.25675938533359266,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049963920749976,
      "loss": 5.6451,
      "mean_token_accuracy": 0.15661601424217225,
      "num_tokens": 5735886.0,
      "step": 3300
    },
    {
      "entropy": 5.874257373809814,
      "epoch": 0.25714841470531025,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996376368696721,
      "loss": 5.6872,
      "mean_token_accuracy": 0.1504574626684189,
      "num_tokens": 5744656.0,
      "step": 3305
    },
    {
      "entropy": 5.962090301513672,
      "epoch": 0.25753744407702783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996360628310656,
      "loss": 5.8558,
      "mean_token_accuracy": 0.14705744832754136,
      "num_tokens": 5753026.0,
      "step": 3310
    },
    {
      "entropy": 5.944463777542114,
      "epoch": 0.25792647344874536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996344853839644,
      "loss": 5.7887,
      "mean_token_accuracy": 0.14836156964302064,
      "num_tokens": 5762771.0,
      "step": 3315
    },
    {
      "entropy": 5.880813026428223,
      "epoch": 0.25831550282046295,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996329045283924,
      "loss": 5.7789,
      "mean_token_accuracy": 0.15456028431653976,
      "num_tokens": 5771151.0,
      "step": 3320
    },
    {
      "entropy": 5.913249826431274,
      "epoch": 0.25870453219218054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996313202643737,
      "loss": 5.6723,
      "mean_token_accuracy": 0.15277076810598372,
      "num_tokens": 5780648.0,
      "step": 3325
    },
    {
      "entropy": 5.894395017623902,
      "epoch": 0.25909356156389807,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996297325919321,
      "loss": 5.6637,
      "mean_token_accuracy": 0.1506694659590721,
      "num_tokens": 5789698.0,
      "step": 3330
    },
    {
      "entropy": 5.952166795730591,
      "epoch": 0.25948259093561565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996281415110919,
      "loss": 5.8531,
      "mean_token_accuracy": 0.14342217296361923,
      "num_tokens": 5799218.0,
      "step": 3335
    },
    {
      "entropy": 5.8528892993927,
      "epoch": 0.2598716203073332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996265470218773,
      "loss": 5.7224,
      "mean_token_accuracy": 0.14743948727846146,
      "num_tokens": 5808455.0,
      "step": 3340
    },
    {
      "entropy": 5.93385009765625,
      "epoch": 0.26026064967905077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996249491243122,
      "loss": 5.7444,
      "mean_token_accuracy": 0.15022578909993173,
      "num_tokens": 5816464.0,
      "step": 3345
    },
    {
      "entropy": 5.881568384170532,
      "epoch": 0.26064967905076836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996233478184211,
      "loss": 5.7044,
      "mean_token_accuracy": 0.1526622548699379,
      "num_tokens": 5825188.0,
      "step": 3350
    },
    {
      "entropy": 5.923000812530518,
      "epoch": 0.2610387084224859,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996217431042282,
      "loss": 5.7652,
      "mean_token_accuracy": 0.15034854263067246,
      "num_tokens": 5834150.0,
      "step": 3355
    },
    {
      "entropy": 5.867080879211426,
      "epoch": 0.2614277377942035,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499620134981758,
      "loss": 5.6124,
      "mean_token_accuracy": 0.1616185039281845,
      "num_tokens": 5842801.0,
      "step": 3360
    },
    {
      "entropy": 5.886935424804688,
      "epoch": 0.261816767165921,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996185234510346,
      "loss": 5.6495,
      "mean_token_accuracy": 0.16229631155729293,
      "num_tokens": 5851547.0,
      "step": 3365
    },
    {
      "entropy": 5.927301025390625,
      "epoch": 0.2622057965376386,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996169085120828,
      "loss": 5.7961,
      "mean_token_accuracy": 0.14882034361362456,
      "num_tokens": 5860430.0,
      "step": 3370
    },
    {
      "entropy": 5.860882091522217,
      "epoch": 0.2625948259093562,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996152901649268,
      "loss": 5.6934,
      "mean_token_accuracy": 0.15090493410825728,
      "num_tokens": 5869648.0,
      "step": 3375
    },
    {
      "entropy": 6.013001251220703,
      "epoch": 0.2629838552810737,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996136684095913,
      "loss": 5.705,
      "mean_token_accuracy": 0.15475423485040665,
      "num_tokens": 5878606.0,
      "step": 3380
    },
    {
      "entropy": 5.845886325836181,
      "epoch": 0.2633728846527913,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996120432461009,
      "loss": 5.7717,
      "mean_token_accuracy": 0.15084365159273147,
      "num_tokens": 5886772.0,
      "step": 3385
    },
    {
      "entropy": 5.840115022659302,
      "epoch": 0.2637619140245088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996104146744804,
      "loss": 5.6822,
      "mean_token_accuracy": 0.15241987705230714,
      "num_tokens": 5895007.0,
      "step": 3390
    },
    {
      "entropy": 6.0143967628479,
      "epoch": 0.2641509433962264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996087826947541,
      "loss": 5.8116,
      "mean_token_accuracy": 0.14449280947446824,
      "num_tokens": 5904249.0,
      "step": 3395
    },
    {
      "entropy": 5.896013498306274,
      "epoch": 0.264539972767944,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996071473069471,
      "loss": 5.7678,
      "mean_token_accuracy": 0.14941729232668877,
      "num_tokens": 5911416.0,
      "step": 3400
    },
    {
      "entropy": 5.942146158218383,
      "epoch": 0.26492900213966153,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996055085110842,
      "loss": 5.8126,
      "mean_token_accuracy": 0.14736545830965042,
      "num_tokens": 5920370.0,
      "step": 3405
    },
    {
      "entropy": 5.916384363174439,
      "epoch": 0.2653180315113791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996038663071902,
      "loss": 5.7647,
      "mean_token_accuracy": 0.1517030656337738,
      "num_tokens": 5928884.0,
      "step": 3410
    },
    {
      "entropy": 5.919314765930176,
      "epoch": 0.26570706088309665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996022206952901,
      "loss": 5.7914,
      "mean_token_accuracy": 0.1425839588046074,
      "num_tokens": 5938456.0,
      "step": 3415
    },
    {
      "entropy": 6.024834537506104,
      "epoch": 0.26609609025481423,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996005716754086,
      "loss": 5.8121,
      "mean_token_accuracy": 0.14940155595541,
      "num_tokens": 5947157.0,
      "step": 3420
    },
    {
      "entropy": 5.928007030487061,
      "epoch": 0.2664851196265318,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499598919247571,
      "loss": 5.6607,
      "mean_token_accuracy": 0.15864564031362532,
      "num_tokens": 5956084.0,
      "step": 3425
    },
    {
      "entropy": 5.8320718765258786,
      "epoch": 0.26687414899824935,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995972634118023,
      "loss": 5.7006,
      "mean_token_accuracy": 0.15306968986988068,
      "num_tokens": 5964782.0,
      "step": 3430
    },
    {
      "entropy": 5.997024011611939,
      "epoch": 0.26726317836996694,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995956041681277,
      "loss": 5.926,
      "mean_token_accuracy": 0.13784579634666444,
      "num_tokens": 5974013.0,
      "step": 3435
    },
    {
      "entropy": 6.002984857559204,
      "epoch": 0.2676522077416845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995939415165722,
      "loss": 5.7924,
      "mean_token_accuracy": 0.14788678735494615,
      "num_tokens": 5983351.0,
      "step": 3440
    },
    {
      "entropy": 5.827401399612427,
      "epoch": 0.26804123711340205,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995922754571611,
      "loss": 5.6321,
      "mean_token_accuracy": 0.15837901681661606,
      "num_tokens": 5992084.0,
      "step": 3445
    },
    {
      "entropy": 5.769678592681885,
      "epoch": 0.26843026648511964,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995906059899197,
      "loss": 5.6461,
      "mean_token_accuracy": 0.15735748410224915,
      "num_tokens": 6001043.0,
      "step": 3450
    },
    {
      "entropy": 5.843425035476685,
      "epoch": 0.26881929585683717,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995889331148733,
      "loss": 5.6494,
      "mean_token_accuracy": 0.1511452838778496,
      "num_tokens": 6009620.0,
      "step": 3455
    },
    {
      "entropy": 5.837132358551026,
      "epoch": 0.26920832522855476,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995872568320474,
      "loss": 5.6575,
      "mean_token_accuracy": 0.15076996386051178,
      "num_tokens": 6018278.0,
      "step": 3460
    },
    {
      "entropy": 5.828963851928711,
      "epoch": 0.26959735460027234,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995855771414673,
      "loss": 5.707,
      "mean_token_accuracy": 0.15559014976024627,
      "num_tokens": 6026741.0,
      "step": 3465
    },
    {
      "entropy": 5.974634027481079,
      "epoch": 0.2699863839719899,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995838940431585,
      "loss": 5.734,
      "mean_token_accuracy": 0.15174644887447358,
      "num_tokens": 6035370.0,
      "step": 3470
    },
    {
      "entropy": 5.874317741394043,
      "epoch": 0.27037541334370746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995822075371466,
      "loss": 5.6988,
      "mean_token_accuracy": 0.15232974737882615,
      "num_tokens": 6044372.0,
      "step": 3475
    },
    {
      "entropy": 5.841263675689698,
      "epoch": 0.270764442715425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995805176234571,
      "loss": 5.719,
      "mean_token_accuracy": 0.15163205564022064,
      "num_tokens": 6052940.0,
      "step": 3480
    },
    {
      "entropy": 5.955672836303711,
      "epoch": 0.2711534720871426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995788243021158,
      "loss": 5.6464,
      "mean_token_accuracy": 0.1543900951743126,
      "num_tokens": 6062048.0,
      "step": 3485
    },
    {
      "entropy": 5.821571111679077,
      "epoch": 0.27154250145886016,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995771275731483,
      "loss": 5.6629,
      "mean_token_accuracy": 0.15446282178163528,
      "num_tokens": 6071478.0,
      "step": 3490
    },
    {
      "entropy": 5.770304489135742,
      "epoch": 0.2719315308305777,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995754274365802,
      "loss": 5.6095,
      "mean_token_accuracy": 0.1592847764492035,
      "num_tokens": 6080495.0,
      "step": 3495
    },
    {
      "entropy": 5.9197368144989015,
      "epoch": 0.2723205602022953,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995737238924376,
      "loss": 5.7657,
      "mean_token_accuracy": 0.14560676813125611,
      "num_tokens": 6090141.0,
      "step": 3500
    },
    {
      "entropy": 5.924958324432373,
      "epoch": 0.2727095895740128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995720169407461,
      "loss": 5.7348,
      "mean_token_accuracy": 0.14742055982351304,
      "num_tokens": 6099215.0,
      "step": 3505
    },
    {
      "entropy": 5.9483616828918455,
      "epoch": 0.2730986189457304,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995703065815317,
      "loss": 5.81,
      "mean_token_accuracy": 0.1458706870675087,
      "num_tokens": 6108802.0,
      "step": 3510
    },
    {
      "entropy": 5.9493061065673825,
      "epoch": 0.273487648317448,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995685928148203,
      "loss": 5.7931,
      "mean_token_accuracy": 0.15170633271336556,
      "num_tokens": 6117486.0,
      "step": 3515
    },
    {
      "entropy": 5.855750560760498,
      "epoch": 0.2738766776891655,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499566875640638,
      "loss": 5.6264,
      "mean_token_accuracy": 0.16193919628858566,
      "num_tokens": 6125248.0,
      "step": 3520
    },
    {
      "entropy": 5.890281057357788,
      "epoch": 0.2742657070608831,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995651550590108,
      "loss": 5.6826,
      "mean_token_accuracy": 0.1505290001630783,
      "num_tokens": 6134831.0,
      "step": 3525
    },
    {
      "entropy": 5.8330237865448,
      "epoch": 0.2746547364326007,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995634310699647,
      "loss": 5.7553,
      "mean_token_accuracy": 0.14910198077559472,
      "num_tokens": 6143248.0,
      "step": 3530
    },
    {
      "entropy": 5.999197769165039,
      "epoch": 0.2750437658043182,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995617036735261,
      "loss": 5.7598,
      "mean_token_accuracy": 0.14741272926330568,
      "num_tokens": 6151768.0,
      "step": 3535
    },
    {
      "entropy": 5.901557064056396,
      "epoch": 0.2754327951760358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995599728697211,
      "loss": 5.6402,
      "mean_token_accuracy": 0.15886410027742387,
      "num_tokens": 6160091.0,
      "step": 3540
    },
    {
      "entropy": 5.8741230964660645,
      "epoch": 0.27582182454775334,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995582386585759,
      "loss": 5.7691,
      "mean_token_accuracy": 0.14154399782419205,
      "num_tokens": 6169440.0,
      "step": 3545
    },
    {
      "entropy": 5.888131856918335,
      "epoch": 0.2762108539194709,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995565010401167,
      "loss": 5.7,
      "mean_token_accuracy": 0.15860531777143477,
      "num_tokens": 6177604.0,
      "step": 3550
    },
    {
      "entropy": 5.918958330154419,
      "epoch": 0.2765998832911885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995547600143702,
      "loss": 5.7376,
      "mean_token_accuracy": 0.1480049178004265,
      "num_tokens": 6186252.0,
      "step": 3555
    },
    {
      "entropy": 5.8245594024658205,
      "epoch": 0.27698891266290604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995530155813625,
      "loss": 5.673,
      "mean_token_accuracy": 0.15742629319429396,
      "num_tokens": 6194641.0,
      "step": 3560
    },
    {
      "entropy": 5.962858867645264,
      "epoch": 0.2773779420346236,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995512677411203,
      "loss": 5.8113,
      "mean_token_accuracy": 0.15001470893621444,
      "num_tokens": 6203925.0,
      "step": 3565
    },
    {
      "entropy": 5.938074159622192,
      "epoch": 0.27776697140634116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995495164936698,
      "loss": 5.6963,
      "mean_token_accuracy": 0.1461140424013138,
      "num_tokens": 6213001.0,
      "step": 3570
    },
    {
      "entropy": 5.865621948242188,
      "epoch": 0.27815600077805874,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995477618390381,
      "loss": 5.719,
      "mean_token_accuracy": 0.14504099935293197,
      "num_tokens": 6221754.0,
      "step": 3575
    },
    {
      "entropy": 5.820681524276734,
      "epoch": 0.27854503014977633,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995460037772513,
      "loss": 5.6025,
      "mean_token_accuracy": 0.15609883069992064,
      "num_tokens": 6229792.0,
      "step": 3580
    },
    {
      "entropy": 5.836638498306274,
      "epoch": 0.27893405952149386,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995442423083365,
      "loss": 5.692,
      "mean_token_accuracy": 0.15051042586565017,
      "num_tokens": 6238839.0,
      "step": 3585
    },
    {
      "entropy": 5.886203908920288,
      "epoch": 0.27932308889321145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995424774323201,
      "loss": 5.6063,
      "mean_token_accuracy": 0.15578331500291825,
      "num_tokens": 6246939.0,
      "step": 3590
    },
    {
      "entropy": 5.932497596740722,
      "epoch": 0.279712118264929,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499540709149229,
      "loss": 5.7214,
      "mean_token_accuracy": 0.15089464485645293,
      "num_tokens": 6255858.0,
      "step": 3595
    },
    {
      "entropy": 5.822576379776001,
      "epoch": 0.28010114763664656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995389374590901,
      "loss": 5.6563,
      "mean_token_accuracy": 0.1534334197640419,
      "num_tokens": 6265137.0,
      "step": 3600
    },
    {
      "entropy": 5.808225059509278,
      "epoch": 0.28049017700836415,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995371623619301,
      "loss": 5.6681,
      "mean_token_accuracy": 0.15250518172979355,
      "num_tokens": 6273286.0,
      "step": 3605
    },
    {
      "entropy": 5.808995914459229,
      "epoch": 0.2808792063800817,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499535383857776,
      "loss": 5.6894,
      "mean_token_accuracy": 0.14724815487861634,
      "num_tokens": 6282477.0,
      "step": 3610
    },
    {
      "entropy": 5.928357744216919,
      "epoch": 0.28126823575179927,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499533601946655,
      "loss": 5.7126,
      "mean_token_accuracy": 0.15294745415449143,
      "num_tokens": 6291756.0,
      "step": 3615
    },
    {
      "entropy": 5.898259210586548,
      "epoch": 0.2816572651235168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995318166285938,
      "loss": 5.798,
      "mean_token_accuracy": 0.14731436520814895,
      "num_tokens": 6301093.0,
      "step": 3620
    },
    {
      "entropy": 5.861028814315796,
      "epoch": 0.2820462944952344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995300279036199,
      "loss": 5.6083,
      "mean_token_accuracy": 0.1591477006673813,
      "num_tokens": 6310379.0,
      "step": 3625
    },
    {
      "entropy": 5.835546207427979,
      "epoch": 0.28243532386695197,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049952823577176,
      "loss": 5.6365,
      "mean_token_accuracy": 0.1520713210105896,
      "num_tokens": 6318983.0,
      "step": 3630
    },
    {
      "entropy": 5.810934829711914,
      "epoch": 0.2828243532386695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995264402330416,
      "loss": 5.6283,
      "mean_token_accuracy": 0.15358254462480544,
      "num_tokens": 6327816.0,
      "step": 3635
    },
    {
      "entropy": 5.81642746925354,
      "epoch": 0.2832133826103871,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995246412874919,
      "loss": 5.7195,
      "mean_token_accuracy": 0.14997470676898955,
      "num_tokens": 6336693.0,
      "step": 3640
    },
    {
      "entropy": 5.894272804260254,
      "epoch": 0.2836024119821047,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499522838935138,
      "loss": 5.654,
      "mean_token_accuracy": 0.15607268661260604,
      "num_tokens": 6345456.0,
      "step": 3645
    },
    {
      "entropy": 5.843640661239624,
      "epoch": 0.2839914413538222,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995210331760074,
      "loss": 5.6625,
      "mean_token_accuracy": 0.15668386667966844,
      "num_tokens": 6353584.0,
      "step": 3650
    },
    {
      "entropy": 5.790026140213013,
      "epoch": 0.2843804707255398,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995192240101275,
      "loss": 5.6837,
      "mean_token_accuracy": 0.15414109826087952,
      "num_tokens": 6362258.0,
      "step": 3655
    },
    {
      "entropy": 5.861711740493774,
      "epoch": 0.2847695000972573,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995174114375258,
      "loss": 5.5981,
      "mean_token_accuracy": 0.1548761434853077,
      "num_tokens": 6371373.0,
      "step": 3660
    },
    {
      "entropy": 5.858789396286011,
      "epoch": 0.2851585294689749,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995155954582297,
      "loss": 5.6692,
      "mean_token_accuracy": 0.15796693712472915,
      "num_tokens": 6380431.0,
      "step": 3665
    },
    {
      "entropy": 5.9123118877410885,
      "epoch": 0.2855475588406925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995137760722668,
      "loss": 5.8339,
      "mean_token_accuracy": 0.1537206545472145,
      "num_tokens": 6389305.0,
      "step": 3670
    },
    {
      "entropy": 5.910440301895141,
      "epoch": 0.28593658821241,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995119532796646,
      "loss": 5.6487,
      "mean_token_accuracy": 0.16165258288383483,
      "num_tokens": 6397065.0,
      "step": 3675
    },
    {
      "entropy": 5.873527336120605,
      "epoch": 0.2863256175841276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499510127080451,
      "loss": 5.7825,
      "mean_token_accuracy": 0.14846334755420684,
      "num_tokens": 6407053.0,
      "step": 3680
    },
    {
      "entropy": 5.942921543121338,
      "epoch": 0.28671464695584514,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995082974746535,
      "loss": 5.7547,
      "mean_token_accuracy": 0.15207315981388092,
      "num_tokens": 6415165.0,
      "step": 3685
    },
    {
      "entropy": 5.920436811447144,
      "epoch": 0.28710367632756273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995064644622999,
      "loss": 5.7195,
      "mean_token_accuracy": 0.15774022042751312,
      "num_tokens": 6423646.0,
      "step": 3690
    },
    {
      "entropy": 5.878801870346069,
      "epoch": 0.2874927056992803,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995046280434181,
      "loss": 5.7397,
      "mean_token_accuracy": 0.15159983560442924,
      "num_tokens": 6432134.0,
      "step": 3695
    },
    {
      "entropy": 5.874917507171631,
      "epoch": 0.28788173507099785,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995027882180358,
      "loss": 5.6316,
      "mean_token_accuracy": 0.15305832773447037,
      "num_tokens": 6440456.0,
      "step": 3700
    },
    {
      "entropy": 5.871138191223144,
      "epoch": 0.28827076444271543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995009449861812,
      "loss": 5.7239,
      "mean_token_accuracy": 0.14909188747406005,
      "num_tokens": 6448952.0,
      "step": 3705
    },
    {
      "entropy": 5.826206254959106,
      "epoch": 0.28865979381443296,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499499098347882,
      "loss": 5.5762,
      "mean_token_accuracy": 0.16490093022584915,
      "num_tokens": 6457225.0,
      "step": 3710
    },
    {
      "entropy": 5.855531024932861,
      "epoch": 0.28904882318615055,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994972483031662,
      "loss": 5.7151,
      "mean_token_accuracy": 0.15100376009941102,
      "num_tokens": 6465450.0,
      "step": 3715
    },
    {
      "entropy": 5.889287090301513,
      "epoch": 0.28943785255786814,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499495394852062,
      "loss": 5.7457,
      "mean_token_accuracy": 0.14744992554187775,
      "num_tokens": 6473736.0,
      "step": 3720
    },
    {
      "entropy": 5.866313505172729,
      "epoch": 0.28982688192958567,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994935379945977,
      "loss": 5.6836,
      "mean_token_accuracy": 0.1602819576859474,
      "num_tokens": 6482092.0,
      "step": 3725
    },
    {
      "entropy": 5.882090663909912,
      "epoch": 0.29021591130130325,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994916777308012,
      "loss": 5.689,
      "mean_token_accuracy": 0.15607897639274598,
      "num_tokens": 6490704.0,
      "step": 3730
    },
    {
      "entropy": 5.948546314239502,
      "epoch": 0.29060494067302084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994898140607007,
      "loss": 5.6464,
      "mean_token_accuracy": 0.15812035351991655,
      "num_tokens": 6499634.0,
      "step": 3735
    },
    {
      "entropy": 5.787464141845703,
      "epoch": 0.29099397004473837,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994879469843247,
      "loss": 5.7928,
      "mean_token_accuracy": 0.14925210997462274,
      "num_tokens": 6509261.0,
      "step": 3740
    },
    {
      "entropy": 5.8902843475341795,
      "epoch": 0.29138299941645596,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994860765017013,
      "loss": 5.6551,
      "mean_token_accuracy": 0.15220843628048897,
      "num_tokens": 6517542.0,
      "step": 3745
    },
    {
      "entropy": 5.859212160110474,
      "epoch": 0.2917720287881735,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499484202612859,
      "loss": 5.6222,
      "mean_token_accuracy": 0.15824688225984573,
      "num_tokens": 6525945.0,
      "step": 3750
    },
    {
      "entropy": 5.767728281021118,
      "epoch": 0.2921610581598911,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994823253178264,
      "loss": 5.6484,
      "mean_token_accuracy": 0.15464899092912673,
      "num_tokens": 6534744.0,
      "step": 3755
    },
    {
      "entropy": 5.840261554718017,
      "epoch": 0.29255008753160866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994804446166317,
      "loss": 5.608,
      "mean_token_accuracy": 0.16365491300821305,
      "num_tokens": 6542640.0,
      "step": 3760
    },
    {
      "entropy": 5.834035968780517,
      "epoch": 0.2929391169033262,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994785605093035,
      "loss": 5.6865,
      "mean_token_accuracy": 0.15791385620832443,
      "num_tokens": 6551292.0,
      "step": 3765
    },
    {
      "entropy": 5.875267362594604,
      "epoch": 0.2933281462750438,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994766729958705,
      "loss": 5.7566,
      "mean_token_accuracy": 0.15267640203237534,
      "num_tokens": 6560760.0,
      "step": 3770
    },
    {
      "entropy": 5.959283542633057,
      "epoch": 0.2937171756467613,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994747820763612,
      "loss": 5.7681,
      "mean_token_accuracy": 0.14788751378655435,
      "num_tokens": 6570618.0,
      "step": 3775
    },
    {
      "entropy": 5.8953528881073,
      "epoch": 0.2941062050184789,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994728877508046,
      "loss": 5.6382,
      "mean_token_accuracy": 0.15356697887182236,
      "num_tokens": 6579484.0,
      "step": 3780
    },
    {
      "entropy": 5.866683626174927,
      "epoch": 0.2944952343901965,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994709900192289,
      "loss": 5.685,
      "mean_token_accuracy": 0.15228400826454164,
      "num_tokens": 6588531.0,
      "step": 3785
    },
    {
      "entropy": 5.870986700057983,
      "epoch": 0.294884263761914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994690888816635,
      "loss": 5.7556,
      "mean_token_accuracy": 0.15102168917655945,
      "num_tokens": 6596862.0,
      "step": 3790
    },
    {
      "entropy": 5.837507677078247,
      "epoch": 0.2952732931336316,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994671843381367,
      "loss": 5.6501,
      "mean_token_accuracy": 0.15472938120365143,
      "num_tokens": 6605594.0,
      "step": 3795
    },
    {
      "entropy": 5.914930200576782,
      "epoch": 0.29566232250534913,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994652763886778,
      "loss": 5.7024,
      "mean_token_accuracy": 0.15655220597982406,
      "num_tokens": 6613971.0,
      "step": 3800
    },
    {
      "entropy": 5.853056573867798,
      "epoch": 0.2960513518770667,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994633650333155,
      "loss": 5.7753,
      "mean_token_accuracy": 0.14286859333515167,
      "num_tokens": 6623116.0,
      "step": 3805
    },
    {
      "entropy": 5.893979835510254,
      "epoch": 0.2964403812487843,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994614502720792,
      "loss": 5.6422,
      "mean_token_accuracy": 0.15770183503627777,
      "num_tokens": 6631910.0,
      "step": 3810
    },
    {
      "entropy": 5.836010408401489,
      "epoch": 0.29682941062050183,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994595321049974,
      "loss": 5.6728,
      "mean_token_accuracy": 0.16161404103040694,
      "num_tokens": 6640882.0,
      "step": 3815
    },
    {
      "entropy": 5.913960695266724,
      "epoch": 0.2972184399922194,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994576105320994,
      "loss": 5.7609,
      "mean_token_accuracy": 0.15119977593421935,
      "num_tokens": 6649665.0,
      "step": 3820
    },
    {
      "entropy": 5.831139135360718,
      "epoch": 0.29760746936393695,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994556855534146,
      "loss": 5.5195,
      "mean_token_accuracy": 0.16653952449560167,
      "num_tokens": 6658725.0,
      "step": 3825
    },
    {
      "entropy": 5.727788639068604,
      "epoch": 0.29799649873565454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994537571689719,
      "loss": 5.5859,
      "mean_token_accuracy": 0.1621174305677414,
      "num_tokens": 6667403.0,
      "step": 3830
    },
    {
      "entropy": 5.896328401565552,
      "epoch": 0.2983855281073721,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994518253788007,
      "loss": 5.6751,
      "mean_token_accuracy": 0.14980594515800477,
      "num_tokens": 6676673.0,
      "step": 3835
    },
    {
      "entropy": 5.914979743957519,
      "epoch": 0.29877455747908965,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994498901829304,
      "loss": 5.6809,
      "mean_token_accuracy": 0.14734249114990233,
      "num_tokens": 6686212.0,
      "step": 3840
    },
    {
      "entropy": 5.779385614395141,
      "epoch": 0.29916358685080724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994479515813903,
      "loss": 5.6692,
      "mean_token_accuracy": 0.15411499589681626,
      "num_tokens": 6694812.0,
      "step": 3845
    },
    {
      "entropy": 5.871626520156861,
      "epoch": 0.2995526162225248,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994460095742096,
      "loss": 5.644,
      "mean_token_accuracy": 0.15434655100107192,
      "num_tokens": 6703439.0,
      "step": 3850
    },
    {
      "entropy": 5.924510478973389,
      "epoch": 0.29994164559424236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499444064161418,
      "loss": 5.6178,
      "mean_token_accuracy": 0.15686172991991043,
      "num_tokens": 6711465.0,
      "step": 3855
    },
    {
      "entropy": 5.810085678100586,
      "epoch": 0.30033067496595994,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499442115343045,
      "loss": 5.6527,
      "mean_token_accuracy": 0.15281544029712676,
      "num_tokens": 6720526.0,
      "step": 3860
    },
    {
      "entropy": 5.945318651199341,
      "epoch": 0.3007197043376775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994401631191202,
      "loss": 5.8317,
      "mean_token_accuracy": 0.144597065448761,
      "num_tokens": 6728696.0,
      "step": 3865
    },
    {
      "entropy": 5.861696434020996,
      "epoch": 0.30110873370939506,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994382074896731,
      "loss": 5.6847,
      "mean_token_accuracy": 0.15377849191427231,
      "num_tokens": 6737955.0,
      "step": 3870
    },
    {
      "entropy": 5.861507368087769,
      "epoch": 0.30149776308111265,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994362484547335,
      "loss": 5.7587,
      "mean_token_accuracy": 0.14718077778816224,
      "num_tokens": 6746413.0,
      "step": 3875
    },
    {
      "entropy": 5.825746250152588,
      "epoch": 0.3018867924528302,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994342860143309,
      "loss": 5.6077,
      "mean_token_accuracy": 0.15988363325595856,
      "num_tokens": 6754341.0,
      "step": 3880
    },
    {
      "entropy": 5.749278116226196,
      "epoch": 0.30227582182454776,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994323201684953,
      "loss": 5.4758,
      "mean_token_accuracy": 0.16771773397922515,
      "num_tokens": 6762623.0,
      "step": 3885
    },
    {
      "entropy": 5.770840740203857,
      "epoch": 0.3026648511962653,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994303509172566,
      "loss": 5.6348,
      "mean_token_accuracy": 0.15647213906049728,
      "num_tokens": 6771270.0,
      "step": 3890
    },
    {
      "entropy": 5.861431550979614,
      "epoch": 0.3030538805679829,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994283782606444,
      "loss": 5.7011,
      "mean_token_accuracy": 0.15365937054157258,
      "num_tokens": 6780292.0,
      "step": 3895
    },
    {
      "entropy": 5.876373672485352,
      "epoch": 0.30344290993970047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994264021986888,
      "loss": 5.5906,
      "mean_token_accuracy": 0.15858400166034697,
      "num_tokens": 6788580.0,
      "step": 3900
    },
    {
      "entropy": 5.7274853706359865,
      "epoch": 0.303831939311418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994244227314197,
      "loss": 5.5919,
      "mean_token_accuracy": 0.1546170249581337,
      "num_tokens": 6796801.0,
      "step": 3905
    },
    {
      "entropy": 5.832391262054443,
      "epoch": 0.3042209686831356,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994224398588672,
      "loss": 5.5681,
      "mean_token_accuracy": 0.15856126993894576,
      "num_tokens": 6805630.0,
      "step": 3910
    },
    {
      "entropy": 5.711365795135498,
      "epoch": 0.3046099980548531,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994204535810615,
      "loss": 5.5667,
      "mean_token_accuracy": 0.1602591648697853,
      "num_tokens": 6814316.0,
      "step": 3915
    },
    {
      "entropy": 5.901001930236816,
      "epoch": 0.3049990274265707,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994184638980326,
      "loss": 5.7496,
      "mean_token_accuracy": 0.14583035409450532,
      "num_tokens": 6823224.0,
      "step": 3920
    },
    {
      "entropy": 5.900492429733276,
      "epoch": 0.3053880567982883,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994164708098107,
      "loss": 5.7687,
      "mean_token_accuracy": 0.14870935082435607,
      "num_tokens": 6832104.0,
      "step": 3925
    },
    {
      "entropy": 5.882292079925537,
      "epoch": 0.3057770861700058,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499414474316426,
      "loss": 5.6663,
      "mean_token_accuracy": 0.15271729081869126,
      "num_tokens": 6841208.0,
      "step": 3930
    },
    {
      "entropy": 5.716450595855713,
      "epoch": 0.3061661155417234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994124744179088,
      "loss": 5.4771,
      "mean_token_accuracy": 0.1640233539044857,
      "num_tokens": 6849622.0,
      "step": 3935
    },
    {
      "entropy": 5.802888536453247,
      "epoch": 0.306555144913441,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994104711142896,
      "loss": 5.667,
      "mean_token_accuracy": 0.1617858126759529,
      "num_tokens": 6857355.0,
      "step": 3940
    },
    {
      "entropy": 5.862309074401855,
      "epoch": 0.3069441742851585,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994084644055986,
      "loss": 5.6882,
      "mean_token_accuracy": 0.16067038625478744,
      "num_tokens": 6866173.0,
      "step": 3945
    },
    {
      "entropy": 5.831095552444458,
      "epoch": 0.3073332036568761,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994064542918664,
      "loss": 5.6897,
      "mean_token_accuracy": 0.14933566004037857,
      "num_tokens": 6874438.0,
      "step": 3950
    },
    {
      "entropy": 5.8412987232208256,
      "epoch": 0.30772223302859364,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994044407731235,
      "loss": 5.6868,
      "mean_token_accuracy": 0.1510934755206108,
      "num_tokens": 6882790.0,
      "step": 3955
    },
    {
      "entropy": 5.82961597442627,
      "epoch": 0.3081112624003112,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994024238494002,
      "loss": 5.7089,
      "mean_token_accuracy": 0.14658345580101012,
      "num_tokens": 6891375.0,
      "step": 3960
    },
    {
      "entropy": 5.759390592575073,
      "epoch": 0.3085002917720288,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994004035207274,
      "loss": 5.6294,
      "mean_token_accuracy": 0.15769621059298516,
      "num_tokens": 6899510.0,
      "step": 3965
    },
    {
      "entropy": 5.903188896179199,
      "epoch": 0.30888932114374634,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993983797871356,
      "loss": 5.6657,
      "mean_token_accuracy": 0.16147721111774443,
      "num_tokens": 6908311.0,
      "step": 3970
    },
    {
      "entropy": 5.7780393123626705,
      "epoch": 0.30927835051546393,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993963526486555,
      "loss": 5.6269,
      "mean_token_accuracy": 0.15729546844959258,
      "num_tokens": 6917064.0,
      "step": 3975
    },
    {
      "entropy": 5.798044586181641,
      "epoch": 0.30966737988718146,
      "grad_norm": 1.0,
      "learning_rate": 0.000499394322105318,
      "loss": 5.6218,
      "mean_token_accuracy": 0.15905228704214097,
      "num_tokens": 6925705.0,
      "step": 3980
    },
    {
      "entropy": 5.966913938522339,
      "epoch": 0.31005640925889905,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993922881571538,
      "loss": 5.7681,
      "mean_token_accuracy": 0.15229437053203582,
      "num_tokens": 6934232.0,
      "step": 3985
    },
    {
      "entropy": 5.856652116775512,
      "epoch": 0.31044543863061663,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993902508041939,
      "loss": 5.7042,
      "mean_token_accuracy": 0.15216735154390335,
      "num_tokens": 6943143.0,
      "step": 3990
    },
    {
      "entropy": 5.810306692123413,
      "epoch": 0.31083446800233416,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499388210046469,
      "loss": 5.5627,
      "mean_token_accuracy": 0.16006666719913482,
      "num_tokens": 6951613.0,
      "step": 3995
    },
    {
      "entropy": 5.857361268997193,
      "epoch": 0.31122349737405175,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993861658840102,
      "loss": 5.7684,
      "mean_token_accuracy": 0.15314728766679764,
      "num_tokens": 6960531.0,
      "step": 4000
    },
    {
      "entropy": 5.887679243087769,
      "epoch": 0.3116125267457693,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993841183168484,
      "loss": 5.7085,
      "mean_token_accuracy": 0.1506664499640465,
      "num_tokens": 6969211.0,
      "step": 4005
    },
    {
      "entropy": 5.951841020584107,
      "epoch": 0.31200155611748687,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993820673450148,
      "loss": 5.7152,
      "mean_token_accuracy": 0.15019422471523286,
      "num_tokens": 6978414.0,
      "step": 4010
    },
    {
      "entropy": 5.760093832015992,
      "epoch": 0.31239058548920445,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993800129685404,
      "loss": 5.4743,
      "mean_token_accuracy": 0.16152779161930084,
      "num_tokens": 6987350.0,
      "step": 4015
    },
    {
      "entropy": 5.764507198333741,
      "epoch": 0.312779614860922,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993779551874565,
      "loss": 5.6139,
      "mean_token_accuracy": 0.16373926550149917,
      "num_tokens": 6995817.0,
      "step": 4020
    },
    {
      "entropy": 5.849539089202881,
      "epoch": 0.31316864423263957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993758940017943,
      "loss": 5.7027,
      "mean_token_accuracy": 0.15734892040491105,
      "num_tokens": 7004454.0,
      "step": 4025
    },
    {
      "entropy": 5.8294881820678714,
      "epoch": 0.31355767360435716,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499373829411585,
      "loss": 5.6349,
      "mean_token_accuracy": 0.15659935623407364,
      "num_tokens": 7012126.0,
      "step": 4030
    },
    {
      "entropy": 5.804521131515503,
      "epoch": 0.3139467029760747,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049937176141686,
      "loss": 5.7087,
      "mean_token_accuracy": 0.1532567746937275,
      "num_tokens": 7021232.0,
      "step": 4035
    },
    {
      "entropy": 5.915135192871094,
      "epoch": 0.3143357323477923,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993696900176506,
      "loss": 5.6783,
      "mean_token_accuracy": 0.157207453250885,
      "num_tokens": 7029488.0,
      "step": 4040
    },
    {
      "entropy": 5.854561281204224,
      "epoch": 0.3147247617195098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993676152139883,
      "loss": 5.7137,
      "mean_token_accuracy": 0.15398113429546356,
      "num_tokens": 7037323.0,
      "step": 4045
    },
    {
      "entropy": 5.821354532241822,
      "epoch": 0.3151137910912274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993655370059046,
      "loss": 5.6607,
      "mean_token_accuracy": 0.15085252672433852,
      "num_tokens": 7045754.0,
      "step": 4050
    },
    {
      "entropy": 5.828005409240722,
      "epoch": 0.315502820462945,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993634553934309,
      "loss": 5.6072,
      "mean_token_accuracy": 0.15728428959846497,
      "num_tokens": 7054026.0,
      "step": 4055
    },
    {
      "entropy": 5.775190687179565,
      "epoch": 0.3158918498346625,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993613703765989,
      "loss": 5.6563,
      "mean_token_accuracy": 0.1554763376712799,
      "num_tokens": 7063383.0,
      "step": 4060
    },
    {
      "entropy": 5.774190759658813,
      "epoch": 0.3162808792063801,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993592819554402,
      "loss": 5.6079,
      "mean_token_accuracy": 0.15743849724531173,
      "num_tokens": 7071952.0,
      "step": 4065
    },
    {
      "entropy": 5.906689405441284,
      "epoch": 0.3166699085780976,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993571901299866,
      "loss": 5.6592,
      "mean_token_accuracy": 0.15330417454242706,
      "num_tokens": 7080283.0,
      "step": 4070
    },
    {
      "entropy": 5.720599365234375,
      "epoch": 0.3170589379498152,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993550949002697,
      "loss": 5.5985,
      "mean_token_accuracy": 0.1576405480504036,
      "num_tokens": 7088669.0,
      "step": 4075
    },
    {
      "entropy": 5.744790935516358,
      "epoch": 0.3174479673215328,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993529962663213,
      "loss": 5.4941,
      "mean_token_accuracy": 0.16759186685085298,
      "num_tokens": 7097201.0,
      "step": 4080
    },
    {
      "entropy": 5.780264663696289,
      "epoch": 0.31783699669325033,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993508942281732,
      "loss": 5.6673,
      "mean_token_accuracy": 0.1554343432188034,
      "num_tokens": 7106509.0,
      "step": 4085
    },
    {
      "entropy": 5.802187919616699,
      "epoch": 0.3182260260649679,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993487887858575,
      "loss": 5.5852,
      "mean_token_accuracy": 0.15869487822055817,
      "num_tokens": 7114550.0,
      "step": 4090
    },
    {
      "entropy": 5.869193267822266,
      "epoch": 0.31861505543668545,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499346679939406,
      "loss": 5.6906,
      "mean_token_accuracy": 0.1567075952887535,
      "num_tokens": 7122875.0,
      "step": 4095
    },
    {
      "entropy": 5.796186923980713,
      "epoch": 0.31900408480840303,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993445676888507,
      "loss": 5.5196,
      "mean_token_accuracy": 0.15808596462011337,
      "num_tokens": 7132263.0,
      "step": 4100
    },
    {
      "entropy": 5.770897054672242,
      "epoch": 0.3193931141801206,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993424520342236,
      "loss": 5.7299,
      "mean_token_accuracy": 0.1540067285299301,
      "num_tokens": 7141448.0,
      "step": 4105
    },
    {
      "entropy": 5.788070201873779,
      "epoch": 0.31978214355183815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993403329755569,
      "loss": 5.7435,
      "mean_token_accuracy": 0.150648832321167,
      "num_tokens": 7151021.0,
      "step": 4110
    },
    {
      "entropy": 5.9307338237762455,
      "epoch": 0.32017117292355574,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993382105128828,
      "loss": 5.6856,
      "mean_token_accuracy": 0.15276920050382614,
      "num_tokens": 7160103.0,
      "step": 4115
    },
    {
      "entropy": 5.851531457901001,
      "epoch": 0.32056020229527327,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993360846462333,
      "loss": 5.684,
      "mean_token_accuracy": 0.1464504860341549,
      "num_tokens": 7169003.0,
      "step": 4120
    },
    {
      "entropy": 5.868911981582642,
      "epoch": 0.32094923166699085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993339553756408,
      "loss": 5.7073,
      "mean_token_accuracy": 0.15806739404797554,
      "num_tokens": 7177086.0,
      "step": 4125
    },
    {
      "entropy": 5.821856737136841,
      "epoch": 0.32133826103870844,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993318227011378,
      "loss": 5.5928,
      "mean_token_accuracy": 0.1515570119023323,
      "num_tokens": 7185772.0,
      "step": 4130
    },
    {
      "entropy": 5.877510070800781,
      "epoch": 0.32172729041042597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993296866227562,
      "loss": 5.7268,
      "mean_token_accuracy": 0.15518796890974046,
      "num_tokens": 7194394.0,
      "step": 4135
    },
    {
      "entropy": 5.814974594116211,
      "epoch": 0.32211631978214356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993275471405288,
      "loss": 5.5416,
      "mean_token_accuracy": 0.16767876148223876,
      "num_tokens": 7202143.0,
      "step": 4140
    },
    {
      "entropy": 5.752964830398559,
      "epoch": 0.32250534915386114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993254042544879,
      "loss": 5.6563,
      "mean_token_accuracy": 0.15957320481538773,
      "num_tokens": 7210645.0,
      "step": 4145
    },
    {
      "entropy": 5.827418470382691,
      "epoch": 0.3228943785255787,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499323257964666,
      "loss": 5.6102,
      "mean_token_accuracy": 0.15694079846143721,
      "num_tokens": 7218872.0,
      "step": 4150
    },
    {
      "entropy": 5.85628924369812,
      "epoch": 0.32328340789729626,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993211082710957,
      "loss": 5.7646,
      "mean_token_accuracy": 0.15412936508655548,
      "num_tokens": 7226977.0,
      "step": 4155
    },
    {
      "entropy": 5.872501277923584,
      "epoch": 0.3236724372690138,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993189551738097,
      "loss": 5.6297,
      "mean_token_accuracy": 0.15481510162353515,
      "num_tokens": 7235635.0,
      "step": 4160
    },
    {
      "entropy": 5.795789957046509,
      "epoch": 0.3240614666407314,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993167986728405,
      "loss": 5.6128,
      "mean_token_accuracy": 0.15973446071147918,
      "num_tokens": 7244506.0,
      "step": 4165
    },
    {
      "entropy": 5.761918640136718,
      "epoch": 0.32445049601244896,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499314638768221,
      "loss": 5.6392,
      "mean_token_accuracy": 0.15619633570313454,
      "num_tokens": 7252883.0,
      "step": 4170
    },
    {
      "entropy": 5.842187452316284,
      "epoch": 0.3248395253841665,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993124754599839,
      "loss": 5.5918,
      "mean_token_accuracy": 0.15565370470285417,
      "num_tokens": 7261580.0,
      "step": 4175
    },
    {
      "entropy": 5.822684812545776,
      "epoch": 0.3252285547558841,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993103087481619,
      "loss": 5.6841,
      "mean_token_accuracy": 0.15471784621477128,
      "num_tokens": 7270346.0,
      "step": 4180
    },
    {
      "entropy": 5.890071296691895,
      "epoch": 0.3256175841276016,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993081386327881,
      "loss": 5.7574,
      "mean_token_accuracy": 0.15108644515275954,
      "num_tokens": 7279393.0,
      "step": 4185
    },
    {
      "entropy": 5.797246932983398,
      "epoch": 0.3260066134993192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993059651138953,
      "loss": 5.582,
      "mean_token_accuracy": 0.15682794600725175,
      "num_tokens": 7287929.0,
      "step": 4190
    },
    {
      "entropy": 5.854136514663696,
      "epoch": 0.3263956428710368,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993037881915165,
      "loss": 5.6517,
      "mean_token_accuracy": 0.1567692741751671,
      "num_tokens": 7296336.0,
      "step": 4195
    },
    {
      "entropy": 5.7092649936676025,
      "epoch": 0.3267846722427543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993016078656847,
      "loss": 5.5201,
      "mean_token_accuracy": 0.16174272447824478,
      "num_tokens": 7304624.0,
      "step": 4200
    },
    {
      "entropy": 5.767000961303711,
      "epoch": 0.3271737016144719,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499299424136433,
      "loss": 5.6565,
      "mean_token_accuracy": 0.15455140471458434,
      "num_tokens": 7312945.0,
      "step": 4205
    },
    {
      "entropy": 5.832268762588501,
      "epoch": 0.32756273098618943,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992972370037946,
      "loss": 5.5886,
      "mean_token_accuracy": 0.1604722708463669,
      "num_tokens": 7321273.0,
      "step": 4210
    },
    {
      "entropy": 5.838960456848144,
      "epoch": 0.327951760357907,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992950464678026,
      "loss": 5.6783,
      "mean_token_accuracy": 0.15889862924814224,
      "num_tokens": 7330446.0,
      "step": 4215
    },
    {
      "entropy": 5.706913661956787,
      "epoch": 0.3283407897296246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992928525284903,
      "loss": 5.6108,
      "mean_token_accuracy": 0.15890374779701233,
      "num_tokens": 7339282.0,
      "step": 4220
    },
    {
      "entropy": 5.799556303024292,
      "epoch": 0.32872981910134214,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499290655185891,
      "loss": 5.592,
      "mean_token_accuracy": 0.15030699148774146,
      "num_tokens": 7347848.0,
      "step": 4225
    },
    {
      "entropy": 5.7438140392303465,
      "epoch": 0.3291188484730597,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499288454440038,
      "loss": 5.5943,
      "mean_token_accuracy": 0.16247443109750748,
      "num_tokens": 7356678.0,
      "step": 4230
    },
    {
      "entropy": 5.805340433120728,
      "epoch": 0.3295078778447773,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992862502909645,
      "loss": 5.6338,
      "mean_token_accuracy": 0.15882461816072463,
      "num_tokens": 7365411.0,
      "step": 4235
    },
    {
      "entropy": 5.7715576171875,
      "epoch": 0.32989690721649484,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992840427387043,
      "loss": 5.5521,
      "mean_token_accuracy": 0.16899469047784804,
      "num_tokens": 7373635.0,
      "step": 4240
    },
    {
      "entropy": 5.794749927520752,
      "epoch": 0.3302859365882124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992818317832908,
      "loss": 5.6832,
      "mean_token_accuracy": 0.15516672283411026,
      "num_tokens": 7382469.0,
      "step": 4245
    },
    {
      "entropy": 5.831536340713501,
      "epoch": 0.33067496595992996,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992796174247574,
      "loss": 5.6601,
      "mean_token_accuracy": 0.15788168013095855,
      "num_tokens": 7390712.0,
      "step": 4250
    },
    {
      "entropy": 5.89034333229065,
      "epoch": 0.33106399533164754,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992773996631378,
      "loss": 5.5945,
      "mean_token_accuracy": 0.15991327166557312,
      "num_tokens": 7398905.0,
      "step": 4255
    },
    {
      "entropy": 5.814719247817993,
      "epoch": 0.33145302470336513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992751784984656,
      "loss": 5.6791,
      "mean_token_accuracy": 0.15588037967681884,
      "num_tokens": 7407242.0,
      "step": 4260
    },
    {
      "entropy": 5.787463998794555,
      "epoch": 0.33184205407508266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992729539307745,
      "loss": 5.6865,
      "mean_token_accuracy": 0.1521645873785019,
      "num_tokens": 7416489.0,
      "step": 4265
    },
    {
      "entropy": 5.855968618392945,
      "epoch": 0.33223108344680025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992707259600984,
      "loss": 5.6738,
      "mean_token_accuracy": 0.16025874763727188,
      "num_tokens": 7424507.0,
      "step": 4270
    },
    {
      "entropy": 5.7866209030151365,
      "epoch": 0.3326201128185178,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992684945864709,
      "loss": 5.6618,
      "mean_token_accuracy": 0.15009636729955672,
      "num_tokens": 7433343.0,
      "step": 4275
    },
    {
      "entropy": 5.760308694839478,
      "epoch": 0.33300914219023536,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992662598099261,
      "loss": 5.5958,
      "mean_token_accuracy": 0.15572042614221573,
      "num_tokens": 7441452.0,
      "step": 4280
    },
    {
      "entropy": 5.8265258312225345,
      "epoch": 0.33339817156195295,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992640216304975,
      "loss": 5.6566,
      "mean_token_accuracy": 0.16005273461341857,
      "num_tokens": 7449541.0,
      "step": 4285
    },
    {
      "entropy": 5.807859754562378,
      "epoch": 0.3337872009336705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992617800482196,
      "loss": 5.6753,
      "mean_token_accuracy": 0.14853483065962791,
      "num_tokens": 7458812.0,
      "step": 4290
    },
    {
      "entropy": 5.822836828231812,
      "epoch": 0.33417623030538807,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499259535063126,
      "loss": 5.5967,
      "mean_token_accuracy": 0.15647593289613723,
      "num_tokens": 7467607.0,
      "step": 4295
    },
    {
      "entropy": 5.809892272949218,
      "epoch": 0.3345652596771056,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499257286675251,
      "loss": 5.5287,
      "mean_token_accuracy": 0.15996506363153457,
      "num_tokens": 7476047.0,
      "step": 4300
    },
    {
      "entropy": 5.819800090789795,
      "epoch": 0.3349542890488232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992550348846285,
      "loss": 5.5922,
      "mean_token_accuracy": 0.15836444348096848,
      "num_tokens": 7484496.0,
      "step": 4305
    },
    {
      "entropy": 5.6688683986663815,
      "epoch": 0.33534331842054077,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992527796912928,
      "loss": 5.533,
      "mean_token_accuracy": 0.16347187012434006,
      "num_tokens": 7493161.0,
      "step": 4310
    },
    {
      "entropy": 5.802170896530152,
      "epoch": 0.3357323477922583,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992505210952782,
      "loss": 5.6606,
      "mean_token_accuracy": 0.15477246642112732,
      "num_tokens": 7502122.0,
      "step": 4315
    },
    {
      "entropy": 5.833450222015381,
      "epoch": 0.3361213771639759,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992482590966188,
      "loss": 5.6368,
      "mean_token_accuracy": 0.15850246399641038,
      "num_tokens": 7510369.0,
      "step": 4320
    },
    {
      "entropy": 5.6804886817932125,
      "epoch": 0.3365104065356934,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992459936953489,
      "loss": 5.5165,
      "mean_token_accuracy": 0.1662609502673149,
      "num_tokens": 7519321.0,
      "step": 4325
    },
    {
      "entropy": 5.744986438751221,
      "epoch": 0.336899435907411,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992437248915032,
      "loss": 5.5699,
      "mean_token_accuracy": 0.16194347888231278,
      "num_tokens": 7528196.0,
      "step": 4330
    },
    {
      "entropy": 5.754108524322509,
      "epoch": 0.3372884652791286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992414526851157,
      "loss": 5.621,
      "mean_token_accuracy": 0.15460745245218277,
      "num_tokens": 7536124.0,
      "step": 4335
    },
    {
      "entropy": 5.849008846282959,
      "epoch": 0.3376774946508461,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992391770762212,
      "loss": 5.7256,
      "mean_token_accuracy": 0.1549411028623581,
      "num_tokens": 7544409.0,
      "step": 4340
    },
    {
      "entropy": 5.797446775436401,
      "epoch": 0.3380665240225637,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499236898064854,
      "loss": 5.6126,
      "mean_token_accuracy": 0.16186843663454056,
      "num_tokens": 7552803.0,
      "step": 4345
    },
    {
      "entropy": 5.758720970153808,
      "epoch": 0.3384555533942813,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992346156510488,
      "loss": 5.5929,
      "mean_token_accuracy": 0.15546914786100388,
      "num_tokens": 7561909.0,
      "step": 4350
    },
    {
      "entropy": 5.852308654785157,
      "epoch": 0.3388445827659988,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992323298348403,
      "loss": 5.633,
      "mean_token_accuracy": 0.15416990593075752,
      "num_tokens": 7571759.0,
      "step": 4355
    },
    {
      "entropy": 5.80660228729248,
      "epoch": 0.3392336121377164,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499230040616263,
      "loss": 5.7731,
      "mean_token_accuracy": 0.15868831053376198,
      "num_tokens": 7580010.0,
      "step": 4360
    },
    {
      "entropy": 5.827027750015259,
      "epoch": 0.33962264150943394,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992277479953518,
      "loss": 5.5593,
      "mean_token_accuracy": 0.16013055592775344,
      "num_tokens": 7587841.0,
      "step": 4365
    },
    {
      "entropy": 5.809354829788208,
      "epoch": 0.34001167088115153,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992254519721414,
      "loss": 5.612,
      "mean_token_accuracy": 0.15086305737495423,
      "num_tokens": 7596187.0,
      "step": 4370
    },
    {
      "entropy": 5.674190807342529,
      "epoch": 0.3404007002528691,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992231525466666,
      "loss": 5.5284,
      "mean_token_accuracy": 0.1612797573208809,
      "num_tokens": 7604361.0,
      "step": 4375
    },
    {
      "entropy": 5.826938152313232,
      "epoch": 0.34078972962458665,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992208497189624,
      "loss": 5.618,
      "mean_token_accuracy": 0.1607014164328575,
      "num_tokens": 7613173.0,
      "step": 4380
    },
    {
      "entropy": 5.761354255676269,
      "epoch": 0.34117875899630423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992185434890637,
      "loss": 5.704,
      "mean_token_accuracy": 0.14778897613286973,
      "num_tokens": 7621730.0,
      "step": 4385
    },
    {
      "entropy": 5.850599241256714,
      "epoch": 0.34156778836802176,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992162338570055,
      "loss": 5.6661,
      "mean_token_accuracy": 0.15339594334363937,
      "num_tokens": 7630362.0,
      "step": 4390
    },
    {
      "entropy": 5.847987699508667,
      "epoch": 0.34195681773973935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992139208228227,
      "loss": 5.6283,
      "mean_token_accuracy": 0.15436543077230452,
      "num_tokens": 7639368.0,
      "step": 4395
    },
    {
      "entropy": 5.795064401626587,
      "epoch": 0.34234584711145694,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992116043865506,
      "loss": 5.6495,
      "mean_token_accuracy": 0.15490932613611222,
      "num_tokens": 7648399.0,
      "step": 4400
    },
    {
      "entropy": 5.761918830871582,
      "epoch": 0.34273487648317447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992092845482244,
      "loss": 5.4742,
      "mean_token_accuracy": 0.1693703129887581,
      "num_tokens": 7657298.0,
      "step": 4405
    },
    {
      "entropy": 5.842507696151733,
      "epoch": 0.34312390585489205,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992069613078791,
      "loss": 5.6373,
      "mean_token_accuracy": 0.15960614532232284,
      "num_tokens": 7665328.0,
      "step": 4410
    },
    {
      "entropy": 5.747133731842041,
      "epoch": 0.3435129352266096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992046346655499,
      "loss": 5.5813,
      "mean_token_accuracy": 0.16045911759138107,
      "num_tokens": 7674610.0,
      "step": 4415
    },
    {
      "entropy": 5.783457612991333,
      "epoch": 0.34390196459832717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992023046212723,
      "loss": 5.6131,
      "mean_token_accuracy": 0.15804652571678163,
      "num_tokens": 7682727.0,
      "step": 4420
    },
    {
      "entropy": 5.768875646591186,
      "epoch": 0.34429099397004476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991999711750816,
      "loss": 5.5786,
      "mean_token_accuracy": 0.16525509506464003,
      "num_tokens": 7691547.0,
      "step": 4425
    },
    {
      "entropy": 5.749335050582886,
      "epoch": 0.3446800233417623,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991976343270133,
      "loss": 5.5455,
      "mean_token_accuracy": 0.15550559759140015,
      "num_tokens": 7699987.0,
      "step": 4430
    },
    {
      "entropy": 5.7058038234710695,
      "epoch": 0.3450690527134799,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991952940771026,
      "loss": 5.6031,
      "mean_token_accuracy": 0.1601826310157776,
      "num_tokens": 7708397.0,
      "step": 4435
    },
    {
      "entropy": 5.769835519790649,
      "epoch": 0.34545808208519746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991929504253852,
      "loss": 5.5657,
      "mean_token_accuracy": 0.15692049711942674,
      "num_tokens": 7716795.0,
      "step": 4440
    },
    {
      "entropy": 5.773987054824829,
      "epoch": 0.345847111456915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991906033718966,
      "loss": 5.605,
      "mean_token_accuracy": 0.15687381029129027,
      "num_tokens": 7725663.0,
      "step": 4445
    },
    {
      "entropy": 5.761317873001099,
      "epoch": 0.3462361408286326,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991882529166724,
      "loss": 5.7216,
      "mean_token_accuracy": 0.15945706516504288,
      "num_tokens": 7735175.0,
      "step": 4450
    },
    {
      "entropy": 5.804883766174316,
      "epoch": 0.3466251702003501,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991858990597483,
      "loss": 5.5777,
      "mean_token_accuracy": 0.16473341584205628,
      "num_tokens": 7744041.0,
      "step": 4455
    },
    {
      "entropy": 5.812509918212891,
      "epoch": 0.3470141995720677,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991835418011601,
      "loss": 5.6411,
      "mean_token_accuracy": 0.15453120768070222,
      "num_tokens": 7752168.0,
      "step": 4460
    },
    {
      "entropy": 5.736903429031372,
      "epoch": 0.3474032289437853,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991811811409434,
      "loss": 5.5876,
      "mean_token_accuracy": 0.158350470662117,
      "num_tokens": 7760746.0,
      "step": 4465
    },
    {
      "entropy": 5.779326963424682,
      "epoch": 0.3477922583155028,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991788170791341,
      "loss": 5.572,
      "mean_token_accuracy": 0.16289048194885253,
      "num_tokens": 7769960.0,
      "step": 4470
    },
    {
      "entropy": 5.830358695983887,
      "epoch": 0.3481812876872204,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991764496157682,
      "loss": 5.5927,
      "mean_token_accuracy": 0.16031462252140044,
      "num_tokens": 7778475.0,
      "step": 4475
    },
    {
      "entropy": 5.695111846923828,
      "epoch": 0.34857031705893793,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991740787508814,
      "loss": 5.477,
      "mean_token_accuracy": 0.1656101241707802,
      "num_tokens": 7787590.0,
      "step": 4480
    },
    {
      "entropy": 5.78382716178894,
      "epoch": 0.3489593464306555,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991717044845097,
      "loss": 5.6542,
      "mean_token_accuracy": 0.15307689160108567,
      "num_tokens": 7796841.0,
      "step": 4485
    },
    {
      "entropy": 5.872925281524658,
      "epoch": 0.3493483758023731,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991693268166892,
      "loss": 5.6484,
      "mean_token_accuracy": 0.1591815985739231,
      "num_tokens": 7804889.0,
      "step": 4490
    },
    {
      "entropy": 5.783057069778442,
      "epoch": 0.34973740517409063,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991669457474559,
      "loss": 5.6415,
      "mean_token_accuracy": 0.16229303032159806,
      "num_tokens": 7813185.0,
      "step": 4495
    },
    {
      "entropy": 5.770741891860962,
      "epoch": 0.3501264345458082,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991645612768461,
      "loss": 5.6333,
      "mean_token_accuracy": 0.1533343091607094,
      "num_tokens": 7821630.0,
      "step": 4500
    },
    {
      "entropy": 5.817403364181518,
      "epoch": 0.35051546391752575,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991621734048958,
      "loss": 5.6606,
      "mean_token_accuracy": 0.16137972474098206,
      "num_tokens": 7829941.0,
      "step": 4505
    },
    {
      "entropy": 5.762375736236573,
      "epoch": 0.35090449328924334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991597821316415,
      "loss": 5.6094,
      "mean_token_accuracy": 0.15952375829219817,
      "num_tokens": 7838515.0,
      "step": 4510
    },
    {
      "entropy": 5.862095975875855,
      "epoch": 0.3512935226609609,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991573874571192,
      "loss": 5.6086,
      "mean_token_accuracy": 0.16334310919046402,
      "num_tokens": 7846575.0,
      "step": 4515
    },
    {
      "entropy": 5.700931644439697,
      "epoch": 0.35168255203267845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991549893813653,
      "loss": 5.4983,
      "mean_token_accuracy": 0.16833383440971375,
      "num_tokens": 7854970.0,
      "step": 4520
    },
    {
      "entropy": 5.756206512451172,
      "epoch": 0.35207158140439604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991525879044162,
      "loss": 5.5846,
      "mean_token_accuracy": 0.16156198233366012,
      "num_tokens": 7863293.0,
      "step": 4525
    },
    {
      "entropy": 5.783585739135742,
      "epoch": 0.35246061077611357,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991501830263084,
      "loss": 5.5863,
      "mean_token_accuracy": 0.1635638877749443,
      "num_tokens": 7871029.0,
      "step": 4530
    },
    {
      "entropy": 5.69771523475647,
      "epoch": 0.35284964014783116,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991477747470783,
      "loss": 5.4488,
      "mean_token_accuracy": 0.16581278443336486,
      "num_tokens": 7879797.0,
      "step": 4535
    },
    {
      "entropy": 5.877959823608398,
      "epoch": 0.35323866951954874,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991453630667625,
      "loss": 5.7541,
      "mean_token_accuracy": 0.15457085222005845,
      "num_tokens": 7888600.0,
      "step": 4540
    },
    {
      "entropy": 5.734695291519165,
      "epoch": 0.3536276988912663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991429479853976,
      "loss": 5.5765,
      "mean_token_accuracy": 0.15913326889276505,
      "num_tokens": 7897794.0,
      "step": 4545
    },
    {
      "entropy": 5.809968614578247,
      "epoch": 0.35401672826298386,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991405295030202,
      "loss": 5.5624,
      "mean_token_accuracy": 0.163884000480175,
      "num_tokens": 7905822.0,
      "step": 4550
    },
    {
      "entropy": 5.694532299041748,
      "epoch": 0.35440575763470145,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991381076196671,
      "loss": 5.6405,
      "mean_token_accuracy": 0.16143991500139238,
      "num_tokens": 7914560.0,
      "step": 4555
    },
    {
      "entropy": 5.667612600326538,
      "epoch": 0.354794787006419,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991356823353748,
      "loss": 5.4739,
      "mean_token_accuracy": 0.1684693694114685,
      "num_tokens": 7924058.0,
      "step": 4560
    },
    {
      "entropy": 5.777330732345581,
      "epoch": 0.35518381637813656,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991332536501804,
      "loss": 5.6355,
      "mean_token_accuracy": 0.15628551989793776,
      "num_tokens": 7932979.0,
      "step": 4565
    },
    {
      "entropy": 5.682309246063232,
      "epoch": 0.3555728457498541,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991308215641205,
      "loss": 5.4203,
      "mean_token_accuracy": 0.17264385670423507,
      "num_tokens": 7941357.0,
      "step": 4570
    },
    {
      "entropy": 5.712415027618408,
      "epoch": 0.3559618751215717,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991283860772323,
      "loss": 5.5851,
      "mean_token_accuracy": 0.16359777003526688,
      "num_tokens": 7949997.0,
      "step": 4575
    },
    {
      "entropy": 5.753382682800293,
      "epoch": 0.35635090449328927,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991259471895525,
      "loss": 5.5255,
      "mean_token_accuracy": 0.15907478630542754,
      "num_tokens": 7958246.0,
      "step": 4580
    },
    {
      "entropy": 5.73652753829956,
      "epoch": 0.3567399338650068,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991235049011182,
      "loss": 5.6265,
      "mean_token_accuracy": 0.152801451086998,
      "num_tokens": 7967176.0,
      "step": 4585
    },
    {
      "entropy": 5.758307361602784,
      "epoch": 0.3571289632367244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991210592119663,
      "loss": 5.4735,
      "mean_token_accuracy": 0.16641369760036467,
      "num_tokens": 7974761.0,
      "step": 4590
    },
    {
      "entropy": 5.68664140701294,
      "epoch": 0.3575179926084419,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991186101221342,
      "loss": 5.5484,
      "mean_token_accuracy": 0.1624031826853752,
      "num_tokens": 7983155.0,
      "step": 4595
    },
    {
      "entropy": 5.744742155075073,
      "epoch": 0.3579070219801595,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499116157631659,
      "loss": 5.5798,
      "mean_token_accuracy": 0.15518496334552764,
      "num_tokens": 7991276.0,
      "step": 4600
    },
    {
      "entropy": 5.826719141006469,
      "epoch": 0.3582960513518771,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991137017405777,
      "loss": 5.6925,
      "mean_token_accuracy": 0.15492383688688277,
      "num_tokens": 7999488.0,
      "step": 4605
    },
    {
      "entropy": 5.794916677474975,
      "epoch": 0.3586850807235946,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991112424489277,
      "loss": 5.5557,
      "mean_token_accuracy": 0.16608401238918305,
      "num_tokens": 8008553.0,
      "step": 4610
    },
    {
      "entropy": 5.691841983795166,
      "epoch": 0.3590741100953122,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991087797567462,
      "loss": 5.5454,
      "mean_token_accuracy": 0.15937763601541519,
      "num_tokens": 8017890.0,
      "step": 4615
    },
    {
      "entropy": 5.780169248580933,
      "epoch": 0.35946313946702974,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991063136640709,
      "loss": 5.6974,
      "mean_token_accuracy": 0.14828183278441429,
      "num_tokens": 8025782.0,
      "step": 4620
    },
    {
      "entropy": 5.8135302543640135,
      "epoch": 0.3598521688387473,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991038441709388,
      "loss": 5.5441,
      "mean_token_accuracy": 0.16591402292251586,
      "num_tokens": 8034623.0,
      "step": 4625
    },
    {
      "entropy": 5.700853729248047,
      "epoch": 0.3602411982104649,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991013712773877,
      "loss": 5.5683,
      "mean_token_accuracy": 0.15996278524398805,
      "num_tokens": 8043789.0,
      "step": 4630
    },
    {
      "entropy": 5.704516410827637,
      "epoch": 0.36063022758218244,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990988949834549,
      "loss": 5.5372,
      "mean_token_accuracy": 0.15648136287927628,
      "num_tokens": 8053456.0,
      "step": 4635
    },
    {
      "entropy": 5.711209392547607,
      "epoch": 0.3610192569539,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499096415289178,
      "loss": 5.5038,
      "mean_token_accuracy": 0.16059173345565797,
      "num_tokens": 8061692.0,
      "step": 4640
    },
    {
      "entropy": 5.6331850528717045,
      "epoch": 0.3614082863256176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990939321945948,
      "loss": 5.4713,
      "mean_token_accuracy": 0.1669237121939659,
      "num_tokens": 8069564.0,
      "step": 4645
    },
    {
      "entropy": 5.751643896102905,
      "epoch": 0.36179731569733514,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990914456997427,
      "loss": 5.5421,
      "mean_token_accuracy": 0.15623156130313873,
      "num_tokens": 8078490.0,
      "step": 4650
    },
    {
      "entropy": 5.726134777069092,
      "epoch": 0.36218634506905273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990889558046598,
      "loss": 5.6331,
      "mean_token_accuracy": 0.15515794157981871,
      "num_tokens": 8086858.0,
      "step": 4655
    },
    {
      "entropy": 5.737401628494263,
      "epoch": 0.36257537444077026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990864625093836,
      "loss": 5.5826,
      "mean_token_accuracy": 0.1621963307261467,
      "num_tokens": 8094542.0,
      "step": 4660
    },
    {
      "entropy": 5.820202112197876,
      "epoch": 0.36296440381248785,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990839658139519,
      "loss": 5.6136,
      "mean_token_accuracy": 0.16103239059448243,
      "num_tokens": 8102756.0,
      "step": 4665
    },
    {
      "entropy": 5.823258686065674,
      "epoch": 0.36335343318420543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990814657184028,
      "loss": 5.6456,
      "mean_token_accuracy": 0.1623107150197029,
      "num_tokens": 8111258.0,
      "step": 4670
    },
    {
      "entropy": 5.87000675201416,
      "epoch": 0.36374246255592296,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990789622227741,
      "loss": 5.7553,
      "mean_token_accuracy": 0.14676559418439866,
      "num_tokens": 8119624.0,
      "step": 4675
    },
    {
      "entropy": 5.73598952293396,
      "epoch": 0.36413149192764055,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990764553271038,
      "loss": 5.5433,
      "mean_token_accuracy": 0.16665254384279252,
      "num_tokens": 8128156.0,
      "step": 4680
    },
    {
      "entropy": 5.703827667236328,
      "epoch": 0.3645205212993581,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990739450314299,
      "loss": 5.5895,
      "mean_token_accuracy": 0.15664388835430146,
      "num_tokens": 8136080.0,
      "step": 4685
    },
    {
      "entropy": 5.797719097137451,
      "epoch": 0.36490955067107567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990714313357906,
      "loss": 5.5569,
      "mean_token_accuracy": 0.16621064841747285,
      "num_tokens": 8144742.0,
      "step": 4690
    },
    {
      "entropy": 5.77716474533081,
      "epoch": 0.36529858004279325,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499068914240224,
      "loss": 5.6307,
      "mean_token_accuracy": 0.16047390848398208,
      "num_tokens": 8153377.0,
      "step": 4695
    },
    {
      "entropy": 5.774137258529663,
      "epoch": 0.3656876094145108,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990663937447682,
      "loss": 5.5609,
      "mean_token_accuracy": 0.16464479267597198,
      "num_tokens": 8162169.0,
      "step": 4700
    },
    {
      "entropy": 5.706748533248901,
      "epoch": 0.36607663878622837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990638698494615,
      "loss": 5.5243,
      "mean_token_accuracy": 0.16574703603982927,
      "num_tokens": 8170727.0,
      "step": 4705
    },
    {
      "entropy": 5.653728389739991,
      "epoch": 0.3664656681579459,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990613425543423,
      "loss": 5.4681,
      "mean_token_accuracy": 0.16673722863197327,
      "num_tokens": 8179013.0,
      "step": 4710
    },
    {
      "entropy": 5.645033836364746,
      "epoch": 0.3668546975296635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990588118594487,
      "loss": 5.4449,
      "mean_token_accuracy": 0.17289653271436692,
      "num_tokens": 8187030.0,
      "step": 4715
    },
    {
      "entropy": 5.673836135864258,
      "epoch": 0.3672437269013811,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990562777648194,
      "loss": 5.4308,
      "mean_token_accuracy": 0.1652565523982048,
      "num_tokens": 8195073.0,
      "step": 4720
    },
    {
      "entropy": 5.671401882171631,
      "epoch": 0.3676327562730986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990537402704928,
      "loss": 5.59,
      "mean_token_accuracy": 0.1610963299870491,
      "num_tokens": 8203626.0,
      "step": 4725
    },
    {
      "entropy": 5.764214181900025,
      "epoch": 0.3680217856448162,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990511993765071,
      "loss": 5.5832,
      "mean_token_accuracy": 0.1626249819993973,
      "num_tokens": 8212750.0,
      "step": 4730
    },
    {
      "entropy": 5.845182847976685,
      "epoch": 0.3684108150165337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990486550829011,
      "loss": 5.587,
      "mean_token_accuracy": 0.15327925831079484,
      "num_tokens": 8221542.0,
      "step": 4735
    },
    {
      "entropy": 5.632064485549927,
      "epoch": 0.3687998443882513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990461073897134,
      "loss": 5.502,
      "mean_token_accuracy": 0.16081211492419242,
      "num_tokens": 8230574.0,
      "step": 4740
    },
    {
      "entropy": 5.761065196990967,
      "epoch": 0.3691888737599689,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990435562969827,
      "loss": 5.5967,
      "mean_token_accuracy": 0.15765105038881302,
      "num_tokens": 8239400.0,
      "step": 4745
    },
    {
      "entropy": 5.768282699584961,
      "epoch": 0.3695779031316864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990410018047475,
      "loss": 5.4434,
      "mean_token_accuracy": 0.16521641463041306,
      "num_tokens": 8247537.0,
      "step": 4750
    },
    {
      "entropy": 5.690558004379272,
      "epoch": 0.369966932503404,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990384439130467,
      "loss": 5.5476,
      "mean_token_accuracy": 0.16147579103708268,
      "num_tokens": 8256551.0,
      "step": 4755
    },
    {
      "entropy": 5.780904102325439,
      "epoch": 0.3703559618751216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990358826219192,
      "loss": 5.5914,
      "mean_token_accuracy": 0.1590727299451828,
      "num_tokens": 8265492.0,
      "step": 4760
    },
    {
      "entropy": 5.764974880218506,
      "epoch": 0.37074499124683913,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990333179314038,
      "loss": 5.5202,
      "mean_token_accuracy": 0.16426937580108641,
      "num_tokens": 8273798.0,
      "step": 4765
    },
    {
      "entropy": 5.664842128753662,
      "epoch": 0.3711340206185567,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990307498415393,
      "loss": 5.5,
      "mean_token_accuracy": 0.16806525439023973,
      "num_tokens": 8281998.0,
      "step": 4770
    },
    {
      "entropy": 5.767178678512574,
      "epoch": 0.37152304999027425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990281783523648,
      "loss": 5.6204,
      "mean_token_accuracy": 0.15874530524015426,
      "num_tokens": 8289833.0,
      "step": 4775
    },
    {
      "entropy": 5.6618608951568605,
      "epoch": 0.37191207936199183,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990256034639192,
      "loss": 5.5905,
      "mean_token_accuracy": 0.16262347549200057,
      "num_tokens": 8297562.0,
      "step": 4780
    },
    {
      "entropy": 5.7884622573852536,
      "epoch": 0.3723011087337094,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990230251762418,
      "loss": 5.5953,
      "mean_token_accuracy": 0.15986310467123985,
      "num_tokens": 8305262.0,
      "step": 4785
    },
    {
      "entropy": 5.737649822235108,
      "epoch": 0.37269013810542695,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990204434893713,
      "loss": 5.5056,
      "mean_token_accuracy": 0.16210581809282304,
      "num_tokens": 8313949.0,
      "step": 4790
    },
    {
      "entropy": 5.712923049926758,
      "epoch": 0.37307916747714454,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990178584033474,
      "loss": 5.5322,
      "mean_token_accuracy": 0.16029467433691025,
      "num_tokens": 8322801.0,
      "step": 4795
    },
    {
      "entropy": 5.791052341461182,
      "epoch": 0.37346819684886207,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990152699182089,
      "loss": 5.5386,
      "mean_token_accuracy": 0.16195625811815262,
      "num_tokens": 8331118.0,
      "step": 4800
    },
    {
      "entropy": 5.729126834869385,
      "epoch": 0.37385722622057965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990126780339953,
      "loss": 5.5936,
      "mean_token_accuracy": 0.16018568128347396,
      "num_tokens": 8340498.0,
      "step": 4805
    },
    {
      "entropy": 5.828460788726806,
      "epoch": 0.37424625559229724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990100827507459,
      "loss": 5.7003,
      "mean_token_accuracy": 0.15686605721712113,
      "num_tokens": 8349235.0,
      "step": 4810
    },
    {
      "entropy": 5.707082319259643,
      "epoch": 0.37463528496401477,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990074840684999,
      "loss": 5.608,
      "mean_token_accuracy": 0.16234321445226668,
      "num_tokens": 8358484.0,
      "step": 4815
    },
    {
      "entropy": 5.752330541610718,
      "epoch": 0.37502431433573236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990048819872969,
      "loss": 5.5245,
      "mean_token_accuracy": 0.16454931795597078,
      "num_tokens": 8366894.0,
      "step": 4820
    },
    {
      "entropy": 5.656948280334473,
      "epoch": 0.3754133437074499,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990022765071765,
      "loss": 5.542,
      "mean_token_accuracy": 0.157652348279953,
      "num_tokens": 8375271.0,
      "step": 4825
    },
    {
      "entropy": 5.720467042922974,
      "epoch": 0.3758023730791675,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498999667628178,
      "loss": 5.4817,
      "mean_token_accuracy": 0.16052767187356948,
      "num_tokens": 8383714.0,
      "step": 4830
    },
    {
      "entropy": 5.666319561004639,
      "epoch": 0.37619140245088506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989970553503411,
      "loss": 5.5777,
      "mean_token_accuracy": 0.16042256951332093,
      "num_tokens": 8392022.0,
      "step": 4835
    },
    {
      "entropy": 5.777729034423828,
      "epoch": 0.3765804318226026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989944396737054,
      "loss": 5.5459,
      "mean_token_accuracy": 0.15841610580682755,
      "num_tokens": 8401282.0,
      "step": 4840
    },
    {
      "entropy": 5.724933242797851,
      "epoch": 0.3769694611943202,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989918205983106,
      "loss": 5.6248,
      "mean_token_accuracy": 0.15712009221315384,
      "num_tokens": 8411147.0,
      "step": 4845
    },
    {
      "entropy": 5.8453539371490475,
      "epoch": 0.37735849056603776,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989891981241964,
      "loss": 5.6121,
      "mean_token_accuracy": 0.16100065112113954,
      "num_tokens": 8420434.0,
      "step": 4850
    },
    {
      "entropy": 5.798123455047607,
      "epoch": 0.3777475199377553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989865722514027,
      "loss": 5.592,
      "mean_token_accuracy": 0.16213036924600602,
      "num_tokens": 8430022.0,
      "step": 4855
    },
    {
      "entropy": 5.573984432220459,
      "epoch": 0.3781365493094729,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989839429799692,
      "loss": 5.3627,
      "mean_token_accuracy": 0.1748791977763176,
      "num_tokens": 8438898.0,
      "step": 4860
    },
    {
      "entropy": 5.610987806320191,
      "epoch": 0.3785255786811904,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498981310309936,
      "loss": 5.4613,
      "mean_token_accuracy": 0.16666309237480165,
      "num_tokens": 8446931.0,
      "step": 4865
    },
    {
      "entropy": 5.758994150161743,
      "epoch": 0.378914608052908,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989786742413428,
      "loss": 5.5365,
      "mean_token_accuracy": 0.16107167899608613,
      "num_tokens": 8455515.0,
      "step": 4870
    },
    {
      "entropy": 5.727382040023803,
      "epoch": 0.3793036374246256,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989760347742298,
      "loss": 5.5562,
      "mean_token_accuracy": 0.15914236456155778,
      "num_tokens": 8464135.0,
      "step": 4875
    },
    {
      "entropy": 5.804942274093628,
      "epoch": 0.3796926667963431,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989733919086369,
      "loss": 5.6331,
      "mean_token_accuracy": 0.1614130526781082,
      "num_tokens": 8473222.0,
      "step": 4880
    },
    {
      "entropy": 5.7484547138214115,
      "epoch": 0.3800816961680607,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989707456446043,
      "loss": 5.6203,
      "mean_token_accuracy": 0.16746980100870132,
      "num_tokens": 8481966.0,
      "step": 4885
    },
    {
      "entropy": 5.753399801254273,
      "epoch": 0.38047072553977823,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989680959821721,
      "loss": 5.6236,
      "mean_token_accuracy": 0.1553274318575859,
      "num_tokens": 8491276.0,
      "step": 4890
    },
    {
      "entropy": 5.7599554538726805,
      "epoch": 0.3808597549114958,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989654429213805,
      "loss": 5.5742,
      "mean_token_accuracy": 0.16141808927059173,
      "num_tokens": 8499926.0,
      "step": 4895
    },
    {
      "entropy": 5.728619718551636,
      "epoch": 0.3812487842832134,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989627864622699,
      "loss": 5.4498,
      "mean_token_accuracy": 0.16520923972129822,
      "num_tokens": 8508544.0,
      "step": 4900
    },
    {
      "entropy": 5.790788269042968,
      "epoch": 0.38163781365493094,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989601266048804,
      "loss": 5.5876,
      "mean_token_accuracy": 0.15025516003370284,
      "num_tokens": 8517326.0,
      "step": 4905
    },
    {
      "entropy": 5.688605213165284,
      "epoch": 0.3820268430266485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989574633492525,
      "loss": 5.5421,
      "mean_token_accuracy": 0.16179005801677704,
      "num_tokens": 8526218.0,
      "step": 4910
    },
    {
      "entropy": 5.771336221694947,
      "epoch": 0.38241587239836605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989547966954266,
      "loss": 5.604,
      "mean_token_accuracy": 0.1568916216492653,
      "num_tokens": 8534708.0,
      "step": 4915
    },
    {
      "entropy": 5.659632444381714,
      "epoch": 0.38280490177008364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498952126643443,
      "loss": 5.5321,
      "mean_token_accuracy": 0.1615009605884552,
      "num_tokens": 8543302.0,
      "step": 4920
    },
    {
      "entropy": 5.671322584152222,
      "epoch": 0.3831939311418012,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989494531933424,
      "loss": 5.4603,
      "mean_token_accuracy": 0.16112909615039825,
      "num_tokens": 8552586.0,
      "step": 4925
    },
    {
      "entropy": 5.689733266830444,
      "epoch": 0.38358296051351876,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989467763451652,
      "loss": 5.4357,
      "mean_token_accuracy": 0.16707136631011962,
      "num_tokens": 8561720.0,
      "step": 4930
    },
    {
      "entropy": 5.715486478805542,
      "epoch": 0.38397198988523634,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989440960989523,
      "loss": 5.4248,
      "mean_token_accuracy": 0.16847473233938218,
      "num_tokens": 8570510.0,
      "step": 4935
    },
    {
      "entropy": 5.712751865386963,
      "epoch": 0.3843610192569539,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498941412454744,
      "loss": 5.5677,
      "mean_token_accuracy": 0.15966854095458985,
      "num_tokens": 8579986.0,
      "step": 4940
    },
    {
      "entropy": 5.700868129730225,
      "epoch": 0.38475004862867146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989387254125813,
      "loss": 5.5321,
      "mean_token_accuracy": 0.15932165682315827,
      "num_tokens": 8588507.0,
      "step": 4945
    },
    {
      "entropy": 5.674834966659546,
      "epoch": 0.38513907800038905,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989360349725049,
      "loss": 5.4326,
      "mean_token_accuracy": 0.16699319183826447,
      "num_tokens": 8596651.0,
      "step": 4950
    },
    {
      "entropy": 5.686874198913574,
      "epoch": 0.3855281073721066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989333411345555,
      "loss": 5.4906,
      "mean_token_accuracy": 0.16555775254964827,
      "num_tokens": 8605080.0,
      "step": 4955
    },
    {
      "entropy": 5.640436172485352,
      "epoch": 0.38591713674382416,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989306438987742,
      "loss": 5.4486,
      "mean_token_accuracy": 0.1623488187789917,
      "num_tokens": 8613017.0,
      "step": 4960
    },
    {
      "entropy": 5.726773118972778,
      "epoch": 0.38630616611554175,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989279432652018,
      "loss": 5.6315,
      "mean_token_accuracy": 0.16297589987516403,
      "num_tokens": 8621954.0,
      "step": 4965
    },
    {
      "entropy": 5.774969434738159,
      "epoch": 0.3866951954872593,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989252392338791,
      "loss": 5.577,
      "mean_token_accuracy": 0.16584077328443528,
      "num_tokens": 8630213.0,
      "step": 4970
    },
    {
      "entropy": 5.799734401702881,
      "epoch": 0.38708422485897687,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989225318048475,
      "loss": 5.5768,
      "mean_token_accuracy": 0.16362085342407226,
      "num_tokens": 8638360.0,
      "step": 4975
    },
    {
      "entropy": 5.744360113143921,
      "epoch": 0.3874732542306944,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989198209781478,
      "loss": 5.6298,
      "mean_token_accuracy": 0.1597909837961197,
      "num_tokens": 8647778.0,
      "step": 4980
    },
    {
      "entropy": 5.75740761756897,
      "epoch": 0.387862283602412,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004989171067538211,
      "loss": 5.5454,
      "mean_token_accuracy": 0.15761953890323638,
      "num_tokens": 8655584.0,
      "step": 4985
    },
    {
      "entropy": 5.68055944442749,
      "epoch": 0.38825131297412957,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498914389131909,
      "loss": 5.5936,
      "mean_token_accuracy": 0.1586558610200882,
      "num_tokens": 8664159.0,
      "step": 4990
    },
    {
      "entropy": 5.707244777679444,
      "epoch": 0.3886403423458471,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989116681124523,
      "loss": 5.4699,
      "mean_token_accuracy": 0.16615322679281236,
      "num_tokens": 8672488.0,
      "step": 4995
    },
    {
      "entropy": 5.778514432907104,
      "epoch": 0.3890293717175647,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989089436954924,
      "loss": 5.6278,
      "mean_token_accuracy": 0.15830786675214767,
      "num_tokens": 8681492.0,
      "step": 5000
    },
    {
      "entropy": 5.666163063049316,
      "epoch": 0.3894184010892822,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989062158810706,
      "loss": 5.512,
      "mean_token_accuracy": 0.16376831531524658,
      "num_tokens": 8690453.0,
      "step": 5005
    },
    {
      "entropy": 5.76589002609253,
      "epoch": 0.3898074304609998,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989034846692284,
      "loss": 5.5933,
      "mean_token_accuracy": 0.16137610152363777,
      "num_tokens": 8699349.0,
      "step": 5010
    },
    {
      "entropy": 5.754378890991211,
      "epoch": 0.3901964598327174,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989007500600073,
      "loss": 5.5986,
      "mean_token_accuracy": 0.15995216518640518,
      "num_tokens": 8707512.0,
      "step": 5015
    },
    {
      "entropy": 5.804534530639648,
      "epoch": 0.3905854892044349,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988980120534486,
      "loss": 5.6108,
      "mean_token_accuracy": 0.16278305947780608,
      "num_tokens": 8715847.0,
      "step": 5020
    },
    {
      "entropy": 5.7424171447753904,
      "epoch": 0.3909745185761525,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498895270649594,
      "loss": 5.5112,
      "mean_token_accuracy": 0.16594649404287337,
      "num_tokens": 8725548.0,
      "step": 5025
    },
    {
      "entropy": 5.649247646331787,
      "epoch": 0.39136354794787004,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498892525848485,
      "loss": 5.4847,
      "mean_token_accuracy": 0.1618966743350029,
      "num_tokens": 8733935.0,
      "step": 5030
    },
    {
      "entropy": 5.708878135681152,
      "epoch": 0.3917525773195876,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988897776501633,
      "loss": 5.5927,
      "mean_token_accuracy": 0.16257513165473939,
      "num_tokens": 8743110.0,
      "step": 5035
    },
    {
      "entropy": 5.669124937057495,
      "epoch": 0.3921416066913052,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988870260546705,
      "loss": 5.4006,
      "mean_token_accuracy": 0.16625102162361144,
      "num_tokens": 8751166.0,
      "step": 5040
    },
    {
      "entropy": 5.723137521743775,
      "epoch": 0.39253063606302274,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988842710620486,
      "loss": 5.5993,
      "mean_token_accuracy": 0.16078926771879196,
      "num_tokens": 8759929.0,
      "step": 5045
    },
    {
      "entropy": 5.761736011505127,
      "epoch": 0.39291966543474033,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988815126723391,
      "loss": 5.6206,
      "mean_token_accuracy": 0.15766884982585908,
      "num_tokens": 8768541.0,
      "step": 5050
    },
    {
      "entropy": 5.767947578430176,
      "epoch": 0.3933086948064579,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988787508855841,
      "loss": 5.5097,
      "mean_token_accuracy": 0.1612896963953972,
      "num_tokens": 8776803.0,
      "step": 5055
    },
    {
      "entropy": 5.694341850280762,
      "epoch": 0.39369772417817545,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004988759857018253,
      "loss": 5.5134,
      "mean_token_accuracy": 0.16799097061157225,
      "num_tokens": 8784773.0,
      "step": 5060
    },
    {
      "entropy": 5.798240900039673,
      "epoch": 0.39408675354989303,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988732171211048,
      "loss": 5.5555,
      "mean_token_accuracy": 0.167237688601017,
      "num_tokens": 8793797.0,
      "step": 5065
    },
    {
      "entropy": 5.640912485122681,
      "epoch": 0.39447578292161056,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988704451434644,
      "loss": 5.4366,
      "mean_token_accuracy": 0.1665719836950302,
      "num_tokens": 8802366.0,
      "step": 5070
    },
    {
      "entropy": 5.742262363433838,
      "epoch": 0.39486481229332815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988676697689465,
      "loss": 5.5711,
      "mean_token_accuracy": 0.1577314779162407,
      "num_tokens": 8810975.0,
      "step": 5075
    },
    {
      "entropy": 5.745263528823853,
      "epoch": 0.39525384166504574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498864890997593,
      "loss": 5.605,
      "mean_token_accuracy": 0.15574074983596803,
      "num_tokens": 8819137.0,
      "step": 5080
    },
    {
      "entropy": 5.761233139038086,
      "epoch": 0.39564287103676327,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988621088294461,
      "loss": 5.5953,
      "mean_token_accuracy": 0.15754300504922866,
      "num_tokens": 8827676.0,
      "step": 5085
    },
    {
      "entropy": 5.737605571746826,
      "epoch": 0.39603190040848085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498859323264548,
      "loss": 5.5405,
      "mean_token_accuracy": 0.16511860489845276,
      "num_tokens": 8836005.0,
      "step": 5090
    },
    {
      "entropy": 5.648690223693848,
      "epoch": 0.3964209297801984,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498856534302941,
      "loss": 5.4519,
      "mean_token_accuracy": 0.1706383228302002,
      "num_tokens": 8844153.0,
      "step": 5095
    },
    {
      "entropy": 5.726143455505371,
      "epoch": 0.39680995915191597,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988537419446673,
      "loss": 5.5587,
      "mean_token_accuracy": 0.15934093594551085,
      "num_tokens": 8852520.0,
      "step": 5100
    },
    {
      "entropy": 5.753759431838989,
      "epoch": 0.39719898852363356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988509461897694,
      "loss": 5.3908,
      "mean_token_accuracy": 0.17240907102823258,
      "num_tokens": 8860609.0,
      "step": 5105
    },
    {
      "entropy": 5.67409610748291,
      "epoch": 0.3975880178953511,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988481470382897,
      "loss": 5.568,
      "mean_token_accuracy": 0.1637749418616295,
      "num_tokens": 8869185.0,
      "step": 5110
    },
    {
      "entropy": 5.687503480911255,
      "epoch": 0.3979770472670687,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988453444902708,
      "loss": 5.4433,
      "mean_token_accuracy": 0.15943183600902558,
      "num_tokens": 8877758.0,
      "step": 5115
    },
    {
      "entropy": 5.7267121315002445,
      "epoch": 0.3983660766387862,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988425385457549,
      "loss": 5.6056,
      "mean_token_accuracy": 0.16380086094141005,
      "num_tokens": 8887236.0,
      "step": 5120
    },
    {
      "entropy": 5.747417783737182,
      "epoch": 0.3987551060105038,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988397292047848,
      "loss": 5.5608,
      "mean_token_accuracy": 0.16228044182062148,
      "num_tokens": 8895808.0,
      "step": 5125
    },
    {
      "entropy": 5.716846132278443,
      "epoch": 0.3991441353822214,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988369164674032,
      "loss": 5.5437,
      "mean_token_accuracy": 0.16444529294967652,
      "num_tokens": 8904091.0,
      "step": 5130
    },
    {
      "entropy": 5.635995244979858,
      "epoch": 0.3995331647539389,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988341003336526,
      "loss": 5.3688,
      "mean_token_accuracy": 0.16816658526659012,
      "num_tokens": 8912046.0,
      "step": 5135
    },
    {
      "entropy": 5.73875241279602,
      "epoch": 0.3999221941256565,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988312808035757,
      "loss": 5.5853,
      "mean_token_accuracy": 0.155656835436821,
      "num_tokens": 8921060.0,
      "step": 5140
    },
    {
      "entropy": 5.70735354423523,
      "epoch": 0.400311223497374,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988284578772155,
      "loss": 5.5182,
      "mean_token_accuracy": 0.1615131378173828,
      "num_tokens": 8930757.0,
      "step": 5145
    },
    {
      "entropy": 5.824063968658447,
      "epoch": 0.4007002528690916,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988256315546146,
      "loss": 5.6853,
      "mean_token_accuracy": 0.15802342146635057,
      "num_tokens": 8939544.0,
      "step": 5150
    },
    {
      "entropy": 5.655792140960694,
      "epoch": 0.4010892822408092,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988228018358161,
      "loss": 5.4993,
      "mean_token_accuracy": 0.16780378073453903,
      "num_tokens": 8949347.0,
      "step": 5155
    },
    {
      "entropy": 5.732742404937744,
      "epoch": 0.40147831161252673,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988199687208628,
      "loss": 5.5399,
      "mean_token_accuracy": 0.16157386302947999,
      "num_tokens": 8957697.0,
      "step": 5160
    },
    {
      "entropy": 5.721534585952758,
      "epoch": 0.4018673409842443,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988171322097977,
      "loss": 5.5781,
      "mean_token_accuracy": 0.15993378311395645,
      "num_tokens": 8966107.0,
      "step": 5165
    },
    {
      "entropy": 5.672869491577148,
      "epoch": 0.4022563703559619,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988142923026638,
      "loss": 5.4833,
      "mean_token_accuracy": 0.16639911830425264,
      "num_tokens": 8974128.0,
      "step": 5170
    },
    {
      "entropy": 5.681173467636109,
      "epoch": 0.40264539972767943,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988114489995044,
      "loss": 5.4719,
      "mean_token_accuracy": 0.16482845544815064,
      "num_tokens": 8982564.0,
      "step": 5175
    },
    {
      "entropy": 5.665506649017334,
      "epoch": 0.403034429099397,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988086023003624,
      "loss": 5.5017,
      "mean_token_accuracy": 0.16397453844547272,
      "num_tokens": 8990924.0,
      "step": 5180
    },
    {
      "entropy": 5.812045288085938,
      "epoch": 0.40342345847111455,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988057522052811,
      "loss": 5.6117,
      "mean_token_accuracy": 0.1547340050339699,
      "num_tokens": 8999827.0,
      "step": 5185
    },
    {
      "entropy": 5.705833959579468,
      "epoch": 0.40381248784283214,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988028987143037,
      "loss": 5.4391,
      "mean_token_accuracy": 0.16634710282087325,
      "num_tokens": 9007885.0,
      "step": 5190
    },
    {
      "entropy": 5.701918840408325,
      "epoch": 0.4042015172145497,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988000418274736,
      "loss": 5.5358,
      "mean_token_accuracy": 0.15985646843910217,
      "num_tokens": 9016752.0,
      "step": 5195
    },
    {
      "entropy": 5.734607696533203,
      "epoch": 0.40459054658626725,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987971815448341,
      "loss": 5.5433,
      "mean_token_accuracy": 0.1646374225616455,
      "num_tokens": 9025547.0,
      "step": 5200
    },
    {
      "entropy": 5.75845513343811,
      "epoch": 0.40497957595798484,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987943178664285,
      "loss": 5.5775,
      "mean_token_accuracy": 0.15561427399516106,
      "num_tokens": 9034711.0,
      "step": 5205
    },
    {
      "entropy": 5.74301700592041,
      "epoch": 0.40536860532970237,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987914507923004,
      "loss": 5.4833,
      "mean_token_accuracy": 0.16486086398363115,
      "num_tokens": 9042565.0,
      "step": 5210
    },
    {
      "entropy": 5.646585178375244,
      "epoch": 0.40575763470141996,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987885803224931,
      "loss": 5.5453,
      "mean_token_accuracy": 0.15442499071359633,
      "num_tokens": 9051590.0,
      "step": 5215
    },
    {
      "entropy": 5.764519596099854,
      "epoch": 0.40614666407313754,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987857064570505,
      "loss": 5.625,
      "mean_token_accuracy": 0.15703439116477966,
      "num_tokens": 9060415.0,
      "step": 5220
    },
    {
      "entropy": 5.677373218536377,
      "epoch": 0.4065356934448551,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987828291960158,
      "loss": 5.4125,
      "mean_token_accuracy": 0.17174528390169144,
      "num_tokens": 9069492.0,
      "step": 5225
    },
    {
      "entropy": 5.65934419631958,
      "epoch": 0.40692472281657266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498779948539433,
      "loss": 5.4867,
      "mean_token_accuracy": 0.1642414942383766,
      "num_tokens": 9078183.0,
      "step": 5230
    },
    {
      "entropy": 5.800519037246704,
      "epoch": 0.4073137521882902,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987770644873455,
      "loss": 5.6877,
      "mean_token_accuracy": 0.15678185373544692,
      "num_tokens": 9086661.0,
      "step": 5235
    },
    {
      "entropy": 5.681110000610351,
      "epoch": 0.4077027815600078,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987741770397974,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17047692239284515,
      "num_tokens": 9095449.0,
      "step": 5240
    },
    {
      "entropy": 5.695486783981323,
      "epoch": 0.40809181093172536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004987712861968323,
      "loss": 5.5524,
      "mean_token_accuracy": 0.16191044598817825,
      "num_tokens": 9103347.0,
      "step": 5245
    },
    {
      "entropy": 5.598839664459229,
      "epoch": 0.4084808403034429,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498768391958494,
      "loss": 5.4639,
      "mean_token_accuracy": 0.16898142844438552,
      "num_tokens": 9112382.0,
      "step": 5250
    },
    {
      "entropy": 5.711949586868286,
      "epoch": 0.4088698696751605,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987654943248266,
      "loss": 5.4157,
      "mean_token_accuracy": 0.16580033898353577,
      "num_tokens": 9120628.0,
      "step": 5255
    },
    {
      "entropy": 5.751555395126343,
      "epoch": 0.40925889904687807,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987625932958739,
      "loss": 5.5733,
      "mean_token_accuracy": 0.1564086616039276,
      "num_tokens": 9130280.0,
      "step": 5260
    },
    {
      "entropy": 5.876784896850586,
      "epoch": 0.4096479284185956,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987596888716801,
      "loss": 5.666,
      "mean_token_accuracy": 0.1538259670138359,
      "num_tokens": 9139720.0,
      "step": 5265
    },
    {
      "entropy": 5.685037422180176,
      "epoch": 0.4100369577903132,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987567810522892,
      "loss": 5.6061,
      "mean_token_accuracy": 0.16040026843547822,
      "num_tokens": 9148537.0,
      "step": 5270
    },
    {
      "entropy": 5.728633642196655,
      "epoch": 0.4104259871620307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987538698377451,
      "loss": 5.6672,
      "mean_token_accuracy": 0.16034265160560607,
      "num_tokens": 9156965.0,
      "step": 5275
    },
    {
      "entropy": 5.71022138595581,
      "epoch": 0.4108150165337483,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987509552280924,
      "loss": 5.3884,
      "mean_token_accuracy": 0.17059710174798964,
      "num_tokens": 9164982.0,
      "step": 5280
    },
    {
      "entropy": 5.721287250518799,
      "epoch": 0.4112040459054659,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498748037223375,
      "loss": 5.6145,
      "mean_token_accuracy": 0.15953994691371917,
      "num_tokens": 9173383.0,
      "step": 5285
    },
    {
      "entropy": 5.781473302841187,
      "epoch": 0.4115930752771834,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987451158236372,
      "loss": 5.6148,
      "mean_token_accuracy": 0.1664327308535576,
      "num_tokens": 9182743.0,
      "step": 5290
    },
    {
      "entropy": 5.766506195068359,
      "epoch": 0.411982104648901,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987421910289234,
      "loss": 5.5047,
      "mean_token_accuracy": 0.1667089506983757,
      "num_tokens": 9191770.0,
      "step": 5295
    },
    {
      "entropy": 5.558193063735962,
      "epoch": 0.41237113402061853,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987392628392781,
      "loss": 5.484,
      "mean_token_accuracy": 0.17050291001796722,
      "num_tokens": 9200885.0,
      "step": 5300
    },
    {
      "entropy": 5.802951240539551,
      "epoch": 0.4127601633923361,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987363312547456,
      "loss": 5.5403,
      "mean_token_accuracy": 0.16366390138864517,
      "num_tokens": 9210114.0,
      "step": 5305
    },
    {
      "entropy": 5.761357307434082,
      "epoch": 0.4131491927640537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987333962753703,
      "loss": 5.5125,
      "mean_token_accuracy": 0.1661704123020172,
      "num_tokens": 9218608.0,
      "step": 5310
    },
    {
      "entropy": 5.674144315719604,
      "epoch": 0.41353822213577124,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987304579011967,
      "loss": 5.5575,
      "mean_token_accuracy": 0.1641923874616623,
      "num_tokens": 9227737.0,
      "step": 5315
    },
    {
      "entropy": 5.6867390155792235,
      "epoch": 0.4139272515074888,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987275161322697,
      "loss": 5.4712,
      "mean_token_accuracy": 0.16756489276885986,
      "num_tokens": 9236589.0,
      "step": 5320
    },
    {
      "entropy": 5.700326871871948,
      "epoch": 0.41431628087920636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987245709686337,
      "loss": 5.5047,
      "mean_token_accuracy": 0.16901690065860747,
      "num_tokens": 9245313.0,
      "step": 5325
    },
    {
      "entropy": 5.692599248886109,
      "epoch": 0.41470531025092394,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987216224103334,
      "loss": 5.5438,
      "mean_token_accuracy": 0.1554454781115055,
      "num_tokens": 9254001.0,
      "step": 5330
    },
    {
      "entropy": 5.778351974487305,
      "epoch": 0.41509433962264153,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987186704574136,
      "loss": 5.6031,
      "mean_token_accuracy": 0.15538137555122375,
      "num_tokens": 9262485.0,
      "step": 5335
    },
    {
      "entropy": 5.701541996002197,
      "epoch": 0.41548336899435906,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987157151099189,
      "loss": 5.4901,
      "mean_token_accuracy": 0.1665266901254654,
      "num_tokens": 9271490.0,
      "step": 5340
    },
    {
      "entropy": 5.663458061218262,
      "epoch": 0.41587239836607665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987127563678945,
      "loss": 5.5304,
      "mean_token_accuracy": 0.16415161043405532,
      "num_tokens": 9280595.0,
      "step": 5345
    },
    {
      "entropy": 5.679933214187622,
      "epoch": 0.4162614277377942,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987097942313852,
      "loss": 5.4348,
      "mean_token_accuracy": 0.17057384997606279,
      "num_tokens": 9289260.0,
      "step": 5350
    },
    {
      "entropy": 5.753614616394043,
      "epoch": 0.41665045710951176,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987068287004355,
      "loss": 5.6049,
      "mean_token_accuracy": 0.16246243417263032,
      "num_tokens": 9297842.0,
      "step": 5355
    },
    {
      "entropy": 5.752865648269653,
      "epoch": 0.41703948648122935,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987038597750909,
      "loss": 5.5993,
      "mean_token_accuracy": 0.15823685377836227,
      "num_tokens": 9306533.0,
      "step": 5360
    },
    {
      "entropy": 5.759689044952393,
      "epoch": 0.4174285158529469,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987008874553964,
      "loss": 5.4835,
      "mean_token_accuracy": 0.1640964165329933,
      "num_tokens": 9314534.0,
      "step": 5365
    },
    {
      "entropy": 5.776584053039551,
      "epoch": 0.41781754522466447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986979117413969,
      "loss": 5.6455,
      "mean_token_accuracy": 0.16746927350759505,
      "num_tokens": 9323179.0,
      "step": 5370
    },
    {
      "entropy": 5.696486854553223,
      "epoch": 0.41820657459638205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986949326331376,
      "loss": 5.4419,
      "mean_token_accuracy": 0.1723403036594391,
      "num_tokens": 9331265.0,
      "step": 5375
    },
    {
      "entropy": 5.697726821899414,
      "epoch": 0.4185956039680996,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986919501306638,
      "loss": 5.5006,
      "mean_token_accuracy": 0.15892638862133027,
      "num_tokens": 9340856.0,
      "step": 5380
    },
    {
      "entropy": 5.706587934494019,
      "epoch": 0.41898463333981717,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986889642340207,
      "loss": 5.5348,
      "mean_token_accuracy": 0.15855901688337326,
      "num_tokens": 9350728.0,
      "step": 5385
    },
    {
      "entropy": 5.78882064819336,
      "epoch": 0.4193736627115347,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986859749432536,
      "loss": 5.6364,
      "mean_token_accuracy": 0.15591242015361786,
      "num_tokens": 9360109.0,
      "step": 5390
    },
    {
      "entropy": 5.7880536079406735,
      "epoch": 0.4197626920832523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986829822584078,
      "loss": 5.6202,
      "mean_token_accuracy": 0.16118184924125672,
      "num_tokens": 9369101.0,
      "step": 5395
    },
    {
      "entropy": 5.728350448608398,
      "epoch": 0.4201517214549699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986799861795289,
      "loss": 5.5667,
      "mean_token_accuracy": 0.1637089490890503,
      "num_tokens": 9377686.0,
      "step": 5400
    },
    {
      "entropy": 5.76845417022705,
      "epoch": 0.4205407508266874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004986769867066622,
      "loss": 5.5557,
      "mean_token_accuracy": 0.16162476390600206,
      "num_tokens": 9386439.0,
      "step": 5405
    },
    {
      "entropy": 5.69290337562561,
      "epoch": 0.420929780198405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986739838398532,
      "loss": 5.6102,
      "mean_token_accuracy": 0.15639652460813522,
      "num_tokens": 9395086.0,
      "step": 5410
    },
    {
      "entropy": 5.688079404830932,
      "epoch": 0.4213188095701225,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986709775791475,
      "loss": 5.461,
      "mean_token_accuracy": 0.16996486335992814,
      "num_tokens": 9403762.0,
      "step": 5415
    },
    {
      "entropy": 5.7050965309143065,
      "epoch": 0.4217078389418401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986679679245907,
      "loss": 5.5091,
      "mean_token_accuracy": 0.1694752588868141,
      "num_tokens": 9412386.0,
      "step": 5420
    },
    {
      "entropy": 5.69061541557312,
      "epoch": 0.4220968683135577,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986649548762286,
      "loss": 5.465,
      "mean_token_accuracy": 0.17578306198120117,
      "num_tokens": 9420538.0,
      "step": 5425
    },
    {
      "entropy": 5.612565708160401,
      "epoch": 0.4224858976852752,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986619384341066,
      "loss": 5.4591,
      "mean_token_accuracy": 0.16782598644495011,
      "num_tokens": 9429083.0,
      "step": 5430
    },
    {
      "entropy": 5.706359767913819,
      "epoch": 0.4228749270569928,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986589185982708,
      "loss": 5.558,
      "mean_token_accuracy": 0.1694366976618767,
      "num_tokens": 9436892.0,
      "step": 5435
    },
    {
      "entropy": 5.7010246276855465,
      "epoch": 0.42326395642871034,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986558953687671,
      "loss": 5.5681,
      "mean_token_accuracy": 0.16823252588510512,
      "num_tokens": 9445267.0,
      "step": 5440
    },
    {
      "entropy": 5.742143821716309,
      "epoch": 0.42365298580042793,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986528687456409,
      "loss": 5.5692,
      "mean_token_accuracy": 0.16173406913876534,
      "num_tokens": 9454292.0,
      "step": 5445
    },
    {
      "entropy": 5.722393798828125,
      "epoch": 0.4240420151721455,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986498387289384,
      "loss": 5.4642,
      "mean_token_accuracy": 0.17036706805229188,
      "num_tokens": 9463126.0,
      "step": 5450
    },
    {
      "entropy": 5.705404376983642,
      "epoch": 0.42443104454386305,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986468053187058,
      "loss": 5.5422,
      "mean_token_accuracy": 0.16476629674434662,
      "num_tokens": 9471703.0,
      "step": 5455
    },
    {
      "entropy": 5.6335962295532225,
      "epoch": 0.42482007391558063,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986437685149887,
      "loss": 5.443,
      "mean_token_accuracy": 0.17423020005226136,
      "num_tokens": 9480029.0,
      "step": 5460
    },
    {
      "entropy": 5.7315908908844,
      "epoch": 0.4252091032872982,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986407283178334,
      "loss": 5.5504,
      "mean_token_accuracy": 0.16471220105886458,
      "num_tokens": 9488277.0,
      "step": 5465
    },
    {
      "entropy": 5.745454263687134,
      "epoch": 0.42559813265901575,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004986376847272861,
      "loss": 5.5909,
      "mean_token_accuracy": 0.16612172424793242,
      "num_tokens": 9496096.0,
      "step": 5470
    },
    {
      "entropy": 5.671952915191651,
      "epoch": 0.42598716203073334,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004986346377433929,
      "loss": 5.4945,
      "mean_token_accuracy": 0.16573259979486465,
      "num_tokens": 9504653.0,
      "step": 5475
    },
    {
      "entropy": 5.73224778175354,
      "epoch": 0.42637619140245087,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986315873661999,
      "loss": 5.4778,
      "mean_token_accuracy": 0.16273901611566544,
      "num_tokens": 9513088.0,
      "step": 5480
    },
    {
      "entropy": 5.654418230056763,
      "epoch": 0.42676522077416845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986285335957536,
      "loss": 5.3953,
      "mean_token_accuracy": 0.1652171418070793,
      "num_tokens": 9521714.0,
      "step": 5485
    },
    {
      "entropy": 5.628698778152466,
      "epoch": 0.42715425014588604,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986254764321002,
      "loss": 5.4998,
      "mean_token_accuracy": 0.16910230666399,
      "num_tokens": 9530629.0,
      "step": 5490
    },
    {
      "entropy": 5.6531576156616214,
      "epoch": 0.42754327951760357,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986224158752861,
      "loss": 5.4805,
      "mean_token_accuracy": 0.17005691379308702,
      "num_tokens": 9539293.0,
      "step": 5495
    },
    {
      "entropy": 5.635426998138428,
      "epoch": 0.42793230888932116,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986193519253578,
      "loss": 5.4397,
      "mean_token_accuracy": 0.17028937339782715,
      "num_tokens": 9547628.0,
      "step": 5500
    },
    {
      "entropy": 5.638662195205688,
      "epoch": 0.4283213382610387,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986162845823618,
      "loss": 5.444,
      "mean_token_accuracy": 0.17030109614133834,
      "num_tokens": 9557260.0,
      "step": 5505
    },
    {
      "entropy": 5.762142133712769,
      "epoch": 0.4287103676327563,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986132138463446,
      "loss": 5.5032,
      "mean_token_accuracy": 0.15915174037218094,
      "num_tokens": 9565923.0,
      "step": 5510
    },
    {
      "entropy": 5.671739196777343,
      "epoch": 0.42909939700447386,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986101397173528,
      "loss": 5.4809,
      "mean_token_accuracy": 0.16655878871679305,
      "num_tokens": 9574373.0,
      "step": 5515
    },
    {
      "entropy": 5.72672061920166,
      "epoch": 0.4294884263761914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498607062195433,
      "loss": 5.5809,
      "mean_token_accuracy": 0.16339072734117507,
      "num_tokens": 9583652.0,
      "step": 5520
    },
    {
      "entropy": 5.714191198348999,
      "epoch": 0.429877455747909,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498603981280632,
      "loss": 5.5423,
      "mean_token_accuracy": 0.1653921753168106,
      "num_tokens": 9592625.0,
      "step": 5525
    },
    {
      "entropy": 5.62770266532898,
      "epoch": 0.4302664851196265,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986008969729964,
      "loss": 5.4371,
      "mean_token_accuracy": 0.1678569346666336,
      "num_tokens": 9600934.0,
      "step": 5530
    },
    {
      "entropy": 5.620938158035278,
      "epoch": 0.4306555144913441,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985978092725731,
      "loss": 5.438,
      "mean_token_accuracy": 0.16914118379354476,
      "num_tokens": 9609481.0,
      "step": 5535
    },
    {
      "entropy": 5.648740673065186,
      "epoch": 0.4310445438630617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498594718179409,
      "loss": 5.3986,
      "mean_token_accuracy": 0.17670834958553314,
      "num_tokens": 9618063.0,
      "step": 5540
    },
    {
      "entropy": 5.684520959854126,
      "epoch": 0.4314335732347792,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985916236935508,
      "loss": 5.4971,
      "mean_token_accuracy": 0.16753216981887817,
      "num_tokens": 9626144.0,
      "step": 5545
    },
    {
      "entropy": 5.505285263061523,
      "epoch": 0.4318226026064968,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985885258150458,
      "loss": 5.4478,
      "mean_token_accuracy": 0.16409727483987807,
      "num_tokens": 9635328.0,
      "step": 5550
    },
    {
      "entropy": 5.708357429504394,
      "epoch": 0.43221163197821433,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985854245439408,
      "loss": 5.5821,
      "mean_token_accuracy": 0.15840986222028733,
      "num_tokens": 9644529.0,
      "step": 5555
    },
    {
      "entropy": 5.714259147644043,
      "epoch": 0.4326006613499319,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985823198802827,
      "loss": 5.5619,
      "mean_token_accuracy": 0.16070234626531602,
      "num_tokens": 9653763.0,
      "step": 5560
    },
    {
      "entropy": 5.677303123474121,
      "epoch": 0.4329896907216495,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985792118241188,
      "loss": 5.4124,
      "mean_token_accuracy": 0.16904089003801345,
      "num_tokens": 9661433.0,
      "step": 5565
    },
    {
      "entropy": 5.617642974853515,
      "epoch": 0.43337872009336703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985761003754961,
      "loss": 5.3728,
      "mean_token_accuracy": 0.17089153975248336,
      "num_tokens": 9669688.0,
      "step": 5570
    },
    {
      "entropy": 5.727176284790039,
      "epoch": 0.4337677494650846,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985729855344622,
      "loss": 5.501,
      "mean_token_accuracy": 0.1625591918826103,
      "num_tokens": 9678403.0,
      "step": 5575
    },
    {
      "entropy": 5.720705986022949,
      "epoch": 0.4341567788368022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498569867301064,
      "loss": 5.6101,
      "mean_token_accuracy": 0.15884283632040025,
      "num_tokens": 9687504.0,
      "step": 5580
    },
    {
      "entropy": 5.6991945743560795,
      "epoch": 0.43454580820851973,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985667456753489,
      "loss": 5.4131,
      "mean_token_accuracy": 0.17116089910268784,
      "num_tokens": 9695772.0,
      "step": 5585
    },
    {
      "entropy": 5.674344444274903,
      "epoch": 0.4349348375802373,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985636206573642,
      "loss": 5.5033,
      "mean_token_accuracy": 0.16308315098285675,
      "num_tokens": 9704150.0,
      "step": 5590
    },
    {
      "entropy": 5.590923118591308,
      "epoch": 0.43532386695195485,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985604922471575,
      "loss": 5.3951,
      "mean_token_accuracy": 0.17157694697380066,
      "num_tokens": 9712245.0,
      "step": 5595
    },
    {
      "entropy": 5.654242897033692,
      "epoch": 0.43571289632367244,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985573604447761,
      "loss": 5.5305,
      "mean_token_accuracy": 0.16453358978033067,
      "num_tokens": 9721027.0,
      "step": 5600
    },
    {
      "entropy": 5.711980819702148,
      "epoch": 0.43610192569539,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985542252502676,
      "loss": 5.4444,
      "mean_token_accuracy": 0.16885592490434648,
      "num_tokens": 9728939.0,
      "step": 5605
    },
    {
      "entropy": 5.7844501495361325,
      "epoch": 0.43649095506710756,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985510866636796,
      "loss": 5.5907,
      "mean_token_accuracy": 0.1631284847855568,
      "num_tokens": 9737863.0,
      "step": 5610
    },
    {
      "entropy": 5.698635053634644,
      "epoch": 0.43687998443882514,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985479446850596,
      "loss": 5.5837,
      "mean_token_accuracy": 0.1567772939801216,
      "num_tokens": 9747666.0,
      "step": 5615
    },
    {
      "entropy": 5.684486770629883,
      "epoch": 0.4372690138105427,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985447993144554,
      "loss": 5.4584,
      "mean_token_accuracy": 0.16718821227550507,
      "num_tokens": 9756595.0,
      "step": 5620
    },
    {
      "entropy": 5.712073850631714,
      "epoch": 0.43765804318226026,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985416505519147,
      "loss": 5.4999,
      "mean_token_accuracy": 0.1649112544953823,
      "num_tokens": 9764747.0,
      "step": 5625
    },
    {
      "entropy": 5.652025318145752,
      "epoch": 0.43804707255397785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985384983974853,
      "loss": 5.4463,
      "mean_token_accuracy": 0.16301581412553787,
      "num_tokens": 9772992.0,
      "step": 5630
    },
    {
      "entropy": 5.686571645736694,
      "epoch": 0.4384361019256954,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985353428512148,
      "loss": 5.4258,
      "mean_token_accuracy": 0.1694833070039749,
      "num_tokens": 9781107.0,
      "step": 5635
    },
    {
      "entropy": 5.647443580627441,
      "epoch": 0.43882513129741296,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985321839131514,
      "loss": 5.5052,
      "mean_token_accuracy": 0.16307013034820556,
      "num_tokens": 9790618.0,
      "step": 5640
    },
    {
      "entropy": 5.717783784866333,
      "epoch": 0.4392141606691305,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985290215833428,
      "loss": 5.5156,
      "mean_token_accuracy": 0.16194633543491363,
      "num_tokens": 9800147.0,
      "step": 5645
    },
    {
      "entropy": 5.715721940994262,
      "epoch": 0.4396031900408481,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985258558618372,
      "loss": 5.5646,
      "mean_token_accuracy": 0.16524495482444762,
      "num_tokens": 9809056.0,
      "step": 5650
    },
    {
      "entropy": 5.618696165084839,
      "epoch": 0.43999221941256567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985226867486825,
      "loss": 5.4272,
      "mean_token_accuracy": 0.16408860236406325,
      "num_tokens": 9817188.0,
      "step": 5655
    },
    {
      "entropy": 5.714680242538452,
      "epoch": 0.4403812487842832,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985195142439267,
      "loss": 5.6097,
      "mean_token_accuracy": 0.15858251750469207,
      "num_tokens": 9826904.0,
      "step": 5660
    },
    {
      "entropy": 5.736992454528808,
      "epoch": 0.4407702781560008,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985163383476181,
      "loss": 5.5723,
      "mean_token_accuracy": 0.16299885511398315,
      "num_tokens": 9835937.0,
      "step": 5665
    },
    {
      "entropy": 5.683105325698852,
      "epoch": 0.44115930752771837,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985131590598048,
      "loss": 5.4063,
      "mean_token_accuracy": 0.1738029733300209,
      "num_tokens": 9843695.0,
      "step": 5670
    },
    {
      "entropy": 5.6560663223266605,
      "epoch": 0.4415483368994359,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985099763805352,
      "loss": 5.4945,
      "mean_token_accuracy": 0.16638974398374556,
      "num_tokens": 9852693.0,
      "step": 5675
    },
    {
      "entropy": 5.623991632461548,
      "epoch": 0.4419373662711535,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985067903098574,
      "loss": 5.4829,
      "mean_token_accuracy": 0.16771433353424073,
      "num_tokens": 9860251.0,
      "step": 5680
    },
    {
      "entropy": 5.615719366073608,
      "epoch": 0.442326395642871,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985036008478197,
      "loss": 5.4156,
      "mean_token_accuracy": 0.1630866065621376,
      "num_tokens": 9868319.0,
      "step": 5685
    },
    {
      "entropy": 5.640031862258911,
      "epoch": 0.4427154250145886,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985004079944707,
      "loss": 5.4628,
      "mean_token_accuracy": 0.16223709285259247,
      "num_tokens": 9877119.0,
      "step": 5690
    },
    {
      "entropy": 5.790488576889038,
      "epoch": 0.4431044543863062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984972117498587,
      "loss": 5.5575,
      "mean_token_accuracy": 0.166337251663208,
      "num_tokens": 9886015.0,
      "step": 5695
    },
    {
      "entropy": 5.686664581298828,
      "epoch": 0.4434934837580237,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984940121140323,
      "loss": 5.4812,
      "mean_token_accuracy": 0.16396376192569734,
      "num_tokens": 9894829.0,
      "step": 5700
    },
    {
      "entropy": 5.795527458190918,
      "epoch": 0.4438825131297413,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049849080908704,
      "loss": 5.7308,
      "mean_token_accuracy": 0.15216802060604095,
      "num_tokens": 9904840.0,
      "step": 5705
    },
    {
      "entropy": 5.80095853805542,
      "epoch": 0.44427154250145884,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984876026689302,
      "loss": 5.5701,
      "mean_token_accuracy": 0.15765475034713744,
      "num_tokens": 9913535.0,
      "step": 5710
    },
    {
      "entropy": 5.7146790504455565,
      "epoch": 0.4446605718731764,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498484392859752,
      "loss": 5.4808,
      "mean_token_accuracy": 0.16396805197000502,
      "num_tokens": 9921676.0,
      "step": 5715
    },
    {
      "entropy": 5.652966928482056,
      "epoch": 0.445049601244894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984811796595538,
      "loss": 5.5499,
      "mean_token_accuracy": 0.16289856731891633,
      "num_tokens": 9930321.0,
      "step": 5720
    },
    {
      "entropy": 5.68314380645752,
      "epoch": 0.44543863061661154,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984779630683843,
      "loss": 5.4838,
      "mean_token_accuracy": 0.16582090109586717,
      "num_tokens": 9939324.0,
      "step": 5725
    },
    {
      "entropy": 5.685439920425415,
      "epoch": 0.44582765998832913,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984747430862924,
      "loss": 5.5041,
      "mean_token_accuracy": 0.1700365960597992,
      "num_tokens": 9948603.0,
      "step": 5730
    },
    {
      "entropy": 5.616638231277466,
      "epoch": 0.44621668936004666,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984715197133271,
      "loss": 5.4122,
      "mean_token_accuracy": 0.16867141276597977,
      "num_tokens": 9957165.0,
      "step": 5735
    },
    {
      "entropy": 5.712148952484131,
      "epoch": 0.44660571873176425,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984682929495371,
      "loss": 5.5817,
      "mean_token_accuracy": 0.1616025224328041,
      "num_tokens": 9966064.0,
      "step": 5740
    },
    {
      "entropy": 5.599972915649414,
      "epoch": 0.44699474810348183,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984650627949715,
      "loss": 5.4283,
      "mean_token_accuracy": 0.16223059445619584,
      "num_tokens": 9975146.0,
      "step": 5745
    },
    {
      "entropy": 5.67345404624939,
      "epoch": 0.44738377747519936,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984618292496792,
      "loss": 5.5229,
      "mean_token_accuracy": 0.16647314578294753,
      "num_tokens": 9983725.0,
      "step": 5750
    },
    {
      "entropy": 5.758822822570801,
      "epoch": 0.44777280684691695,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984585923137093,
      "loss": 5.5541,
      "mean_token_accuracy": 0.16265189945697783,
      "num_tokens": 9992627.0,
      "step": 5755
    },
    {
      "entropy": 5.647780513763427,
      "epoch": 0.4481618362186345,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498455351987111,
      "loss": 5.379,
      "mean_token_accuracy": 0.17269414216279982,
      "num_tokens": 10000801.0,
      "step": 5760
    },
    {
      "entropy": 5.711228275299073,
      "epoch": 0.44855086559035207,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984521082699333,
      "loss": 5.5289,
      "mean_token_accuracy": 0.16474660187959672,
      "num_tokens": 10009421.0,
      "step": 5765
    },
    {
      "entropy": 5.624726390838623,
      "epoch": 0.44893989496206965,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984488611622256,
      "loss": 5.5169,
      "mean_token_accuracy": 0.16462063044309616,
      "num_tokens": 10018527.0,
      "step": 5770
    },
    {
      "entropy": 5.656307363510132,
      "epoch": 0.4493289243337872,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498445610664037,
      "loss": 5.4476,
      "mean_token_accuracy": 0.1701345145702362,
      "num_tokens": 10026906.0,
      "step": 5775
    },
    {
      "entropy": 5.647906255722046,
      "epoch": 0.44971795370550477,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498442356775417,
      "loss": 5.4396,
      "mean_token_accuracy": 0.1710012897849083,
      "num_tokens": 10035972.0,
      "step": 5780
    },
    {
      "entropy": 5.691563510894776,
      "epoch": 0.45010698307722236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984390994964148,
      "loss": 5.5341,
      "mean_token_accuracy": 0.15645694434642793,
      "num_tokens": 10044544.0,
      "step": 5785
    },
    {
      "entropy": 5.74356255531311,
      "epoch": 0.4504960124489399,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049843583882708,
      "loss": 5.5208,
      "mean_token_accuracy": 0.16705722212791443,
      "num_tokens": 10052880.0,
      "step": 5790
    },
    {
      "entropy": 5.725162506103516,
      "epoch": 0.4508850418206575,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498432574767462,
      "loss": 5.5795,
      "mean_token_accuracy": 0.1562555268406868,
      "num_tokens": 10062373.0,
      "step": 5795
    },
    {
      "entropy": 5.6625590324401855,
      "epoch": 0.451274071192375,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984293073176103,
      "loss": 5.3862,
      "mean_token_accuracy": 0.17116446793079376,
      "num_tokens": 10070473.0,
      "step": 5800
    },
    {
      "entropy": 5.719627332687378,
      "epoch": 0.4516631005640926,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984260364775744,
      "loss": 5.5189,
      "mean_token_accuracy": 0.16526814699172973,
      "num_tokens": 10079857.0,
      "step": 5805
    },
    {
      "entropy": 5.7111578464508055,
      "epoch": 0.4520521299358102,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498422762247404,
      "loss": 5.4724,
      "mean_token_accuracy": 0.1682709887623787,
      "num_tokens": 10088326.0,
      "step": 5810
    },
    {
      "entropy": 5.700175476074219,
      "epoch": 0.4524411593075277,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984194846271489,
      "loss": 5.4182,
      "mean_token_accuracy": 0.16691567450761796,
      "num_tokens": 10097208.0,
      "step": 5815
    },
    {
      "entropy": 5.6884301662445065,
      "epoch": 0.4528301886792453,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984162036168586,
      "loss": 5.5285,
      "mean_token_accuracy": 0.16202297061681747,
      "num_tokens": 10106057.0,
      "step": 5820
    },
    {
      "entropy": 5.6238037109375,
      "epoch": 0.4532192180509628,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984129192165831,
      "loss": 5.4342,
      "mean_token_accuracy": 0.16660418659448623,
      "num_tokens": 10115384.0,
      "step": 5825
    },
    {
      "entropy": 5.694101476669312,
      "epoch": 0.4536082474226804,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984096314263722,
      "loss": 5.5278,
      "mean_token_accuracy": 0.16025744676589965,
      "num_tokens": 10123958.0,
      "step": 5830
    },
    {
      "entropy": 5.730915689468384,
      "epoch": 0.453997276794398,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984063402462757,
      "loss": 5.4673,
      "mean_token_accuracy": 0.1708914190530777,
      "num_tokens": 10133199.0,
      "step": 5835
    },
    {
      "entropy": 5.665083169937134,
      "epoch": 0.45438630616611553,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984030456763435,
      "loss": 5.4256,
      "mean_token_accuracy": 0.17243075519800186,
      "num_tokens": 10141796.0,
      "step": 5840
    },
    {
      "entropy": 5.650837755203247,
      "epoch": 0.4547753355378331,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983997477166257,
      "loss": 5.4456,
      "mean_token_accuracy": 0.16942461729049682,
      "num_tokens": 10149941.0,
      "step": 5845
    },
    {
      "entropy": 5.680732202529907,
      "epoch": 0.45516436490955064,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983964463671723,
      "loss": 5.5194,
      "mean_token_accuracy": 0.16303732842206956,
      "num_tokens": 10158796.0,
      "step": 5850
    },
    {
      "entropy": 5.688786792755127,
      "epoch": 0.45555339428126823,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983931416280334,
      "loss": 5.4715,
      "mean_token_accuracy": 0.1647087052464485,
      "num_tokens": 10167328.0,
      "step": 5855
    },
    {
      "entropy": 5.686644268035889,
      "epoch": 0.4559424236529858,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983898334992591,
      "loss": 5.5311,
      "mean_token_accuracy": 0.1607556089758873,
      "num_tokens": 10175937.0,
      "step": 5860
    },
    {
      "entropy": 5.683647871017456,
      "epoch": 0.45633145302470335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004983865219808998,
      "loss": 5.483,
      "mean_token_accuracy": 0.15861394256353378,
      "num_tokens": 10184828.0,
      "step": 5865
    },
    {
      "entropy": 5.650877952575684,
      "epoch": 0.45672048239642093,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983832070730055,
      "loss": 5.4828,
      "mean_token_accuracy": 0.17043073177337648,
      "num_tokens": 10193871.0,
      "step": 5870
    },
    {
      "entropy": 5.6467538356781,
      "epoch": 0.4571095117681385,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983798887756265,
      "loss": 5.4837,
      "mean_token_accuracy": 0.16615234166383744,
      "num_tokens": 10202740.0,
      "step": 5875
    },
    {
      "entropy": 5.667803430557251,
      "epoch": 0.45749854113985605,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983765670888133,
      "loss": 5.4391,
      "mean_token_accuracy": 0.16772495806217194,
      "num_tokens": 10210903.0,
      "step": 5880
    },
    {
      "entropy": 5.593382549285889,
      "epoch": 0.45788757051157364,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983732420126163,
      "loss": 5.3811,
      "mean_token_accuracy": 0.17336988300085068,
      "num_tokens": 10219305.0,
      "step": 5885
    },
    {
      "entropy": 5.730053281784057,
      "epoch": 0.45827659988329117,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983699135470858,
      "loss": 5.5407,
      "mean_token_accuracy": 0.15575051307678223,
      "num_tokens": 10228575.0,
      "step": 5890
    },
    {
      "entropy": 5.724256181716919,
      "epoch": 0.45866562925500876,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983665816922723,
      "loss": 5.5317,
      "mean_token_accuracy": 0.1638547196984291,
      "num_tokens": 10237813.0,
      "step": 5895
    },
    {
      "entropy": 5.729646492004394,
      "epoch": 0.45905465862672634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983632464482267,
      "loss": 5.5239,
      "mean_token_accuracy": 0.1646823763847351,
      "num_tokens": 10246606.0,
      "step": 5900
    },
    {
      "entropy": 5.59196286201477,
      "epoch": 0.4594436879984439,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004983599078149991,
      "loss": 5.4672,
      "mean_token_accuracy": 0.1709397628903389,
      "num_tokens": 10254996.0,
      "step": 5905
    },
    {
      "entropy": 5.6242005825042725,
      "epoch": 0.45983271737016146,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983565657926405,
      "loss": 5.532,
      "mean_token_accuracy": 0.16023842841386796,
      "num_tokens": 10263570.0,
      "step": 5910
    },
    {
      "entropy": 5.648565435409546,
      "epoch": 0.460221746741879,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983532203812017,
      "loss": 5.3797,
      "mean_token_accuracy": 0.17313698977231978,
      "num_tokens": 10272237.0,
      "step": 5915
    },
    {
      "entropy": 5.67567548751831,
      "epoch": 0.4606107761135966,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983498715807331,
      "loss": 5.5043,
      "mean_token_accuracy": 0.1646430239081383,
      "num_tokens": 10281215.0,
      "step": 5920
    },
    {
      "entropy": 5.718982076644897,
      "epoch": 0.46099980548531416,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983465193912857,
      "loss": 5.5568,
      "mean_token_accuracy": 0.16212098002433778,
      "num_tokens": 10291054.0,
      "step": 5925
    },
    {
      "entropy": 5.757201385498047,
      "epoch": 0.4613888348570317,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983431638129104,
      "loss": 5.5401,
      "mean_token_accuracy": 0.16678355485200883,
      "num_tokens": 10299847.0,
      "step": 5930
    },
    {
      "entropy": 5.650951480865478,
      "epoch": 0.4617778642287493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983398048456581,
      "loss": 5.4622,
      "mean_token_accuracy": 0.17062322944402694,
      "num_tokens": 10309195.0,
      "step": 5935
    },
    {
      "entropy": 5.680678176879883,
      "epoch": 0.4621668936004668,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983364424895796,
      "loss": 5.4873,
      "mean_token_accuracy": 0.1661696419119835,
      "num_tokens": 10318386.0,
      "step": 5940
    },
    {
      "entropy": 5.628504467010498,
      "epoch": 0.4625559229721844,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983330767447262,
      "loss": 5.3888,
      "mean_token_accuracy": 0.16944532841444016,
      "num_tokens": 10327230.0,
      "step": 5945
    },
    {
      "entropy": 5.61873197555542,
      "epoch": 0.462944952343902,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983297076111489,
      "loss": 5.392,
      "mean_token_accuracy": 0.17210619598627092,
      "num_tokens": 10335617.0,
      "step": 5950
    },
    {
      "entropy": 5.669249582290649,
      "epoch": 0.4633339817156195,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983263350888987,
      "loss": 5.4704,
      "mean_token_accuracy": 0.16894796639680862,
      "num_tokens": 10344302.0,
      "step": 5955
    },
    {
      "entropy": 5.692659473419189,
      "epoch": 0.4637230110873371,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004983229591780268,
      "loss": 5.517,
      "mean_token_accuracy": 0.16277996301651002,
      "num_tokens": 10352490.0,
      "step": 5960
    },
    {
      "entropy": 5.65666241645813,
      "epoch": 0.46411204045905463,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983195798785844,
      "loss": 5.447,
      "mean_token_accuracy": 0.16850241869688035,
      "num_tokens": 10361249.0,
      "step": 5965
    },
    {
      "entropy": 5.705624389648437,
      "epoch": 0.4645010698307722,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983161971906228,
      "loss": 5.5381,
      "mean_token_accuracy": 0.160544715821743,
      "num_tokens": 10370367.0,
      "step": 5970
    },
    {
      "entropy": 5.6782958030700685,
      "epoch": 0.4648900992024898,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983128111141935,
      "loss": 5.5139,
      "mean_token_accuracy": 0.16174531430006028,
      "num_tokens": 10378429.0,
      "step": 5975
    },
    {
      "entropy": 5.658778047561645,
      "epoch": 0.46527912857420733,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983094216493475,
      "loss": 5.4224,
      "mean_token_accuracy": 0.16716012358665466,
      "num_tokens": 10386762.0,
      "step": 5980
    },
    {
      "entropy": 5.639854764938354,
      "epoch": 0.4656681579459249,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983060287961367,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1661702111363411,
      "num_tokens": 10395150.0,
      "step": 5985
    },
    {
      "entropy": 5.6702272415161135,
      "epoch": 0.4660571873176425,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983026325546123,
      "loss": 5.459,
      "mean_token_accuracy": 0.1709236368536949,
      "num_tokens": 10403941.0,
      "step": 5990
    },
    {
      "entropy": 5.759852409362793,
      "epoch": 0.46644621668936004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982992329248257,
      "loss": 5.5393,
      "mean_token_accuracy": 0.15724412649869918,
      "num_tokens": 10413955.0,
      "step": 5995
    },
    {
      "entropy": 5.684827375411987,
      "epoch": 0.4668352460610776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004982958299068289,
      "loss": 5.4739,
      "mean_token_accuracy": 0.1667661651968956,
      "num_tokens": 10421676.0,
      "step": 6000
    },
    {
      "epoch": 0.4668352460610776,
      "eval_entropy": 5.431893091229319,
      "eval_loss": 5.5091681480407715,
      "eval_mean_token_accuracy": 0.17318276160031815,
      "eval_num_tokens": 10421676.0,
      "eval_runtime": 21.796,
      "eval_samples_per_second": 1906.678,
      "eval_steps_per_second": 238.346,
      "step": 6000
    },
    {
      "entropy": 5.608576107025146,
      "epoch": 0.46722427543279516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498292423500673,
      "loss": 5.4398,
      "mean_token_accuracy": 0.17045483589172364,
      "num_tokens": 10430714.0,
      "step": 6005
    },
    {
      "entropy": 5.741376781463623,
      "epoch": 0.46761330480451274,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982890137064102,
      "loss": 5.5471,
      "mean_token_accuracy": 0.16484737694263457,
      "num_tokens": 10439089.0,
      "step": 6010
    },
    {
      "entropy": 5.6508848667144775,
      "epoch": 0.46800233417623033,
      "grad_norm": 1.0,
      "learning_rate": 0.000498285600524092,
      "loss": 5.4615,
      "mean_token_accuracy": 0.16931159645318986,
      "num_tokens": 10447455.0,
      "step": 6015
    },
    {
      "entropy": 5.6349437713623045,
      "epoch": 0.46839136354794786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982821839537701,
      "loss": 5.4678,
      "mean_token_accuracy": 0.16561769992113112,
      "num_tokens": 10455936.0,
      "step": 6020
    },
    {
      "entropy": 5.75211615562439,
      "epoch": 0.46878039291966545,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982787639954966,
      "loss": 5.5923,
      "mean_token_accuracy": 0.15538098365068437,
      "num_tokens": 10465389.0,
      "step": 6025
    },
    {
      "entropy": 5.783642387390136,
      "epoch": 0.469169422291383,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982753406493232,
      "loss": 5.4733,
      "mean_token_accuracy": 0.1611355021595955,
      "num_tokens": 10474709.0,
      "step": 6030
    },
    {
      "entropy": 5.646716070175171,
      "epoch": 0.46955845166310056,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982719139153018,
      "loss": 5.4636,
      "mean_token_accuracy": 0.16188982576131822,
      "num_tokens": 10482972.0,
      "step": 6035
    },
    {
      "entropy": 5.6523096561431885,
      "epoch": 0.46994748103481815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982684837934845,
      "loss": 5.481,
      "mean_token_accuracy": 0.16894741505384445,
      "num_tokens": 10490897.0,
      "step": 6040
    },
    {
      "entropy": 5.619273662567139,
      "epoch": 0.4703365104065357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982650502839234,
      "loss": 5.4739,
      "mean_token_accuracy": 0.16739446222782134,
      "num_tokens": 10499679.0,
      "step": 6045
    },
    {
      "entropy": 5.722547817230224,
      "epoch": 0.47072553977825327,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982616133866705,
      "loss": 5.5225,
      "mean_token_accuracy": 0.1597910165786743,
      "num_tokens": 10508775.0,
      "step": 6050
    },
    {
      "entropy": 5.640441989898681,
      "epoch": 0.4711145691499708,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982581731017779,
      "loss": 5.4174,
      "mean_token_accuracy": 0.17352351695299148,
      "num_tokens": 10517600.0,
      "step": 6055
    },
    {
      "entropy": 5.526639747619629,
      "epoch": 0.4715035985216884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982547294292979,
      "loss": 5.3753,
      "mean_token_accuracy": 0.17222920656204224,
      "num_tokens": 10525674.0,
      "step": 6060
    },
    {
      "entropy": 5.670207786560058,
      "epoch": 0.47189262789340597,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982512823692828,
      "loss": 5.5764,
      "mean_token_accuracy": 0.1645464688539505,
      "num_tokens": 10533819.0,
      "step": 6065
    },
    {
      "entropy": 5.665865468978882,
      "epoch": 0.4722816572651235,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982478319217848,
      "loss": 5.3799,
      "mean_token_accuracy": 0.17236743867397308,
      "num_tokens": 10542901.0,
      "step": 6070
    },
    {
      "entropy": 5.668054580688477,
      "epoch": 0.4726706866368411,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982443780868562,
      "loss": 5.4694,
      "mean_token_accuracy": 0.16793531328439712,
      "num_tokens": 10552388.0,
      "step": 6075
    },
    {
      "entropy": 5.753566884994507,
      "epoch": 0.4730597160085587,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982409208645496,
      "loss": 5.5524,
      "mean_token_accuracy": 0.1613159030675888,
      "num_tokens": 10561676.0,
      "step": 6080
    },
    {
      "entropy": 5.723232460021973,
      "epoch": 0.4734487453802762,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982374602549173,
      "loss": 5.4628,
      "mean_token_accuracy": 0.1703447327017784,
      "num_tokens": 10570037.0,
      "step": 6085
    },
    {
      "entropy": 5.613542985916138,
      "epoch": 0.4738377747519938,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982339962580119,
      "loss": 5.456,
      "mean_token_accuracy": 0.16250336319208145,
      "num_tokens": 10579140.0,
      "step": 6090
    },
    {
      "entropy": 5.643436861038208,
      "epoch": 0.4742268041237113,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982305288738859,
      "loss": 5.5145,
      "mean_token_accuracy": 0.1653794452548027,
      "num_tokens": 10587689.0,
      "step": 6095
    },
    {
      "entropy": 5.695700597763062,
      "epoch": 0.4746158334954289,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982270581025919,
      "loss": 5.393,
      "mean_token_accuracy": 0.17702221274375915,
      "num_tokens": 10596002.0,
      "step": 6100
    },
    {
      "entropy": 5.698622703552246,
      "epoch": 0.4750048628671465,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982235839441826,
      "loss": 5.6104,
      "mean_token_accuracy": 0.1616410955786705,
      "num_tokens": 10605895.0,
      "step": 6105
    },
    {
      "entropy": 5.72393479347229,
      "epoch": 0.475393892238864,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982201063987108,
      "loss": 5.5397,
      "mean_token_accuracy": 0.16163956373929977,
      "num_tokens": 10614641.0,
      "step": 6110
    },
    {
      "entropy": 5.671689176559449,
      "epoch": 0.4757829216105816,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982166254662292,
      "loss": 5.5063,
      "mean_token_accuracy": 0.1643342435359955,
      "num_tokens": 10624381.0,
      "step": 6115
    },
    {
      "entropy": 5.658418941497803,
      "epoch": 0.47617195098229914,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982131411467904,
      "loss": 5.4881,
      "mean_token_accuracy": 0.16424735337495805,
      "num_tokens": 10632493.0,
      "step": 6120
    },
    {
      "entropy": 5.6930074214935305,
      "epoch": 0.4765609803540167,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982096534404476,
      "loss": 5.4895,
      "mean_token_accuracy": 0.15681614726781845,
      "num_tokens": 10641556.0,
      "step": 6125
    },
    {
      "entropy": 5.681613540649414,
      "epoch": 0.4769500097257343,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982061623472535,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16099169477820396,
      "num_tokens": 10649736.0,
      "step": 6130
    },
    {
      "entropy": 5.638624572753907,
      "epoch": 0.47733903909745184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982026678672612,
      "loss": 5.388,
      "mean_token_accuracy": 0.1735039621591568,
      "num_tokens": 10657991.0,
      "step": 6135
    },
    {
      "entropy": 5.603248023986817,
      "epoch": 0.47772806846916943,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981991700005238,
      "loss": 5.444,
      "mean_token_accuracy": 0.1621242195367813,
      "num_tokens": 10667310.0,
      "step": 6140
    },
    {
      "entropy": 5.788144159317016,
      "epoch": 0.47811709784088696,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498195668747094,
      "loss": 5.5048,
      "mean_token_accuracy": 0.16948303282260896,
      "num_tokens": 10675658.0,
      "step": 6145
    },
    {
      "entropy": 5.66076512336731,
      "epoch": 0.47850612721260455,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981921641070254,
      "loss": 5.3826,
      "mean_token_accuracy": 0.177694571018219,
      "num_tokens": 10684016.0,
      "step": 6150
    },
    {
      "entropy": 5.634742164611817,
      "epoch": 0.47889515658432213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981886560803708,
      "loss": 5.5155,
      "mean_token_accuracy": 0.16364313811063766,
      "num_tokens": 10692814.0,
      "step": 6155
    },
    {
      "entropy": 5.703499889373779,
      "epoch": 0.47928418595603967,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981851446671838,
      "loss": 5.5152,
      "mean_token_accuracy": 0.16615862548351287,
      "num_tokens": 10701571.0,
      "step": 6160
    },
    {
      "entropy": 5.7480603694915775,
      "epoch": 0.47967321532775725,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981816298675173,
      "loss": 5.4904,
      "mean_token_accuracy": 0.16190893203020096,
      "num_tokens": 10710569.0,
      "step": 6165
    },
    {
      "entropy": 5.704768991470337,
      "epoch": 0.4800622446994748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981781116814248,
      "loss": 5.53,
      "mean_token_accuracy": 0.16690579205751419,
      "num_tokens": 10719277.0,
      "step": 6170
    },
    {
      "entropy": 5.611009645462036,
      "epoch": 0.48045127407119237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981745901089596,
      "loss": 5.4,
      "mean_token_accuracy": 0.16918960511684417,
      "num_tokens": 10727174.0,
      "step": 6175
    },
    {
      "entropy": 5.668210077285766,
      "epoch": 0.48084030344290996,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981710651501753,
      "loss": 5.5189,
      "mean_token_accuracy": 0.1595567524433136,
      "num_tokens": 10736072.0,
      "step": 6180
    },
    {
      "entropy": 5.771615982055664,
      "epoch": 0.4812293328146275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981675368051254,
      "loss": 5.51,
      "mean_token_accuracy": 0.17003518491983413,
      "num_tokens": 10745375.0,
      "step": 6185
    },
    {
      "entropy": 5.663010454177856,
      "epoch": 0.4816183621863451,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981640050738633,
      "loss": 5.5023,
      "mean_token_accuracy": 0.16472843140363694,
      "num_tokens": 10754049.0,
      "step": 6190
    },
    {
      "entropy": 5.5936798572540285,
      "epoch": 0.48200739155806266,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981604699564426,
      "loss": 5.4438,
      "mean_token_accuracy": 0.16255026012659074,
      "num_tokens": 10763475.0,
      "step": 6195
    },
    {
      "entropy": 5.6747595310211185,
      "epoch": 0.4823964209297802,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981569314529169,
      "loss": 5.5092,
      "mean_token_accuracy": 0.16237585097551346,
      "num_tokens": 10772100.0,
      "step": 6200
    },
    {
      "entropy": 5.661139869689942,
      "epoch": 0.4827854503014978,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049815338956334,
      "loss": 5.4652,
      "mean_token_accuracy": 0.16773047298192978,
      "num_tokens": 10780605.0,
      "step": 6205
    },
    {
      "entropy": 5.699847602844239,
      "epoch": 0.4831744796732153,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981498442877656,
      "loss": 5.4378,
      "mean_token_accuracy": 0.1654672622680664,
      "num_tokens": 10789235.0,
      "step": 6210
    },
    {
      "entropy": 5.676222133636474,
      "epoch": 0.4835635090449329,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981462956262474,
      "loss": 5.4379,
      "mean_token_accuracy": 0.16388014554977418,
      "num_tokens": 10798078.0,
      "step": 6215
    },
    {
      "entropy": 5.701656341552734,
      "epoch": 0.4839525384166505,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981427435788396,
      "loss": 5.476,
      "mean_token_accuracy": 0.16313115358352662,
      "num_tokens": 10807667.0,
      "step": 6220
    },
    {
      "entropy": 5.659240579605102,
      "epoch": 0.484341567788368,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981391881455957,
      "loss": 5.4326,
      "mean_token_accuracy": 0.16373479664325713,
      "num_tokens": 10816574.0,
      "step": 6225
    },
    {
      "entropy": 5.602508211135865,
      "epoch": 0.4847305971600856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981356293265696,
      "loss": 5.4349,
      "mean_token_accuracy": 0.16597452610731125,
      "num_tokens": 10824675.0,
      "step": 6230
    },
    {
      "entropy": 5.649123811721802,
      "epoch": 0.4851196265318031,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981320671218157,
      "loss": 5.3892,
      "mean_token_accuracy": 0.17281295508146285,
      "num_tokens": 10832855.0,
      "step": 6235
    },
    {
      "entropy": 5.552571630477905,
      "epoch": 0.4855086559035207,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981285015313877,
      "loss": 5.2839,
      "mean_token_accuracy": 0.1768725708127022,
      "num_tokens": 10841603.0,
      "step": 6240
    },
    {
      "entropy": 5.722947072982788,
      "epoch": 0.4858976852752383,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981249325553399,
      "loss": 5.6348,
      "mean_token_accuracy": 0.15403055697679519,
      "num_tokens": 10850348.0,
      "step": 6245
    },
    {
      "entropy": 5.717859220504761,
      "epoch": 0.48628671464695583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981213601937264,
      "loss": 5.3645,
      "mean_token_accuracy": 0.17039883732795716,
      "num_tokens": 10858955.0,
      "step": 6250
    },
    {
      "entropy": 5.680239057540893,
      "epoch": 0.4866757440186734,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981177844466013,
      "loss": 5.4683,
      "mean_token_accuracy": 0.16246459633111954,
      "num_tokens": 10867336.0,
      "step": 6255
    },
    {
      "entropy": 5.667668294906616,
      "epoch": 0.48706477339039095,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981142053140192,
      "loss": 5.5273,
      "mean_token_accuracy": 0.1636221796274185,
      "num_tokens": 10877043.0,
      "step": 6260
    },
    {
      "entropy": 5.74675612449646,
      "epoch": 0.48745380276210853,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981106227960339,
      "loss": 5.4924,
      "mean_token_accuracy": 0.16905430257320403,
      "num_tokens": 10885580.0,
      "step": 6265
    },
    {
      "entropy": 5.727813482284546,
      "epoch": 0.4878428321338261,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981070368927001,
      "loss": 5.6018,
      "mean_token_accuracy": 0.1581639528274536,
      "num_tokens": 10894289.0,
      "step": 6270
    },
    {
      "entropy": 5.715327215194702,
      "epoch": 0.48823186150554365,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498103447604072,
      "loss": 5.517,
      "mean_token_accuracy": 0.158392234146595,
      "num_tokens": 10902384.0,
      "step": 6275
    },
    {
      "entropy": 5.6708290576934814,
      "epoch": 0.48862089087726124,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980998549302044,
      "loss": 5.4309,
      "mean_token_accuracy": 0.17125301510095597,
      "num_tokens": 10911160.0,
      "step": 6280
    },
    {
      "entropy": 5.609786701202393,
      "epoch": 0.4890099202489788,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980962588711516,
      "loss": 5.3262,
      "mean_token_accuracy": 0.17744140475988388,
      "num_tokens": 10920066.0,
      "step": 6285
    },
    {
      "entropy": 5.692303276062011,
      "epoch": 0.48939894962069636,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498092659426968,
      "loss": 5.5,
      "mean_token_accuracy": 0.16787974089384078,
      "num_tokens": 10928969.0,
      "step": 6290
    },
    {
      "entropy": 5.7058093547821045,
      "epoch": 0.48978797899241394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980890565977085,
      "loss": 5.5191,
      "mean_token_accuracy": 0.16595033705234527,
      "num_tokens": 10936939.0,
      "step": 6295
    },
    {
      "entropy": 5.649946308135986,
      "epoch": 0.4901770083641315,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980854503834276,
      "loss": 5.4424,
      "mean_token_accuracy": 0.16394996643066406,
      "num_tokens": 10945929.0,
      "step": 6300
    },
    {
      "entropy": 5.677580070495606,
      "epoch": 0.49056603773584906,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980818407841802,
      "loss": 5.4484,
      "mean_token_accuracy": 0.16683558970689774,
      "num_tokens": 10954849.0,
      "step": 6305
    },
    {
      "entropy": 5.665102910995484,
      "epoch": 0.49095506710756665,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004980782278000207,
      "loss": 5.5227,
      "mean_token_accuracy": 0.16242400258779527,
      "num_tokens": 10963800.0,
      "step": 6310
    },
    {
      "entropy": 5.813465976715088,
      "epoch": 0.4913440964792842,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980746114310043,
      "loss": 5.5629,
      "mean_token_accuracy": 0.16588517874479294,
      "num_tokens": 10972477.0,
      "step": 6315
    },
    {
      "entropy": 5.7394904613494875,
      "epoch": 0.49173312585100176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980709916771858,
      "loss": 5.574,
      "mean_token_accuracy": 0.15996729359030723,
      "num_tokens": 10981437.0,
      "step": 6320
    },
    {
      "entropy": 5.64530758857727,
      "epoch": 0.4921221552227193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980673685386198,
      "loss": 5.3954,
      "mean_token_accuracy": 0.1674257069826126,
      "num_tokens": 10989973.0,
      "step": 6325
    },
    {
      "entropy": 5.595485687255859,
      "epoch": 0.4925111845944369,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980637420153618,
      "loss": 5.4387,
      "mean_token_accuracy": 0.1638665169477463,
      "num_tokens": 10998160.0,
      "step": 6330
    },
    {
      "entropy": 5.694740152359008,
      "epoch": 0.49290021396615447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980601121074664,
      "loss": 5.5688,
      "mean_token_accuracy": 0.16035808622837067,
      "num_tokens": 11007127.0,
      "step": 6335
    },
    {
      "entropy": 5.693447637557983,
      "epoch": 0.493289243337872,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980564788149889,
      "loss": 5.4374,
      "mean_token_accuracy": 0.16625568270683289,
      "num_tokens": 11015518.0,
      "step": 6340
    },
    {
      "entropy": 5.559444856643677,
      "epoch": 0.4936782727095896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980528421379844,
      "loss": 5.329,
      "mean_token_accuracy": 0.17834123522043227,
      "num_tokens": 11023739.0,
      "step": 6345
    },
    {
      "entropy": 5.590536737442017,
      "epoch": 0.4940673020813071,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498049202076508,
      "loss": 5.3573,
      "mean_token_accuracy": 0.17453927546739578,
      "num_tokens": 11032034.0,
      "step": 6350
    },
    {
      "entropy": 5.632488489151001,
      "epoch": 0.4944563314530247,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980455586306149,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1736905425786972,
      "num_tokens": 11040319.0,
      "step": 6355
    },
    {
      "entropy": 5.537005805969239,
      "epoch": 0.4948453608247423,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980419118003605,
      "loss": 5.4304,
      "mean_token_accuracy": 0.16703475266695023,
      "num_tokens": 11049459.0,
      "step": 6360
    },
    {
      "entropy": 5.690892553329467,
      "epoch": 0.4952343901964598,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980382615858001,
      "loss": 5.4235,
      "mean_token_accuracy": 0.16820919662714004,
      "num_tokens": 11058402.0,
      "step": 6365
    },
    {
      "entropy": 5.6239392280578615,
      "epoch": 0.4956234195681774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980346079869892,
      "loss": 5.3726,
      "mean_token_accuracy": 0.1734849825501442,
      "num_tokens": 11067430.0,
      "step": 6370
    },
    {
      "entropy": 5.63100643157959,
      "epoch": 0.49601244893989493,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498030951003983,
      "loss": 5.3758,
      "mean_token_accuracy": 0.16949167847633362,
      "num_tokens": 11075594.0,
      "step": 6375
    },
    {
      "entropy": 5.613965845108032,
      "epoch": 0.4964014783116125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980272906368371,
      "loss": 5.3392,
      "mean_token_accuracy": 0.18207351863384247,
      "num_tokens": 11083509.0,
      "step": 6380
    },
    {
      "entropy": 5.684510564804077,
      "epoch": 0.4967905076833301,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980236268856071,
      "loss": 5.4953,
      "mean_token_accuracy": 0.16970266103744508,
      "num_tokens": 11091938.0,
      "step": 6385
    },
    {
      "entropy": 5.662152528762817,
      "epoch": 0.49717953705504764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980199597503487,
      "loss": 5.5107,
      "mean_token_accuracy": 0.1673850014805794,
      "num_tokens": 11099974.0,
      "step": 6390
    },
    {
      "entropy": 5.520812129974365,
      "epoch": 0.4975685664267652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980162892311171,
      "loss": 5.1809,
      "mean_token_accuracy": 0.18767566978931427,
      "num_tokens": 11107944.0,
      "step": 6395
    },
    {
      "entropy": 5.561925792694092,
      "epoch": 0.4979575957984828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980126153279684,
      "loss": 5.4691,
      "mean_token_accuracy": 0.17101271003484725,
      "num_tokens": 11116401.0,
      "step": 6400
    },
    {
      "entropy": 5.680764198303223,
      "epoch": 0.49834662517020034,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980089380409583,
      "loss": 5.4385,
      "mean_token_accuracy": 0.16166648268699646,
      "num_tokens": 11125430.0,
      "step": 6405
    },
    {
      "entropy": 5.714994716644287,
      "epoch": 0.4987356545419179,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980052573701424,
      "loss": 5.5023,
      "mean_token_accuracy": 0.16102435290813447,
      "num_tokens": 11133949.0,
      "step": 6410
    },
    {
      "entropy": 5.671395254135132,
      "epoch": 0.49912468391363546,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980015733155767,
      "loss": 5.44,
      "mean_token_accuracy": 0.17307015955448152,
      "num_tokens": 11141749.0,
      "step": 6415
    },
    {
      "entropy": 5.613661623001098,
      "epoch": 0.49951371328535304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979978858773171,
      "loss": 5.4345,
      "mean_token_accuracy": 0.1723210409283638,
      "num_tokens": 11150618.0,
      "step": 6420
    },
    {
      "entropy": 5.631605768203736,
      "epoch": 0.49990274265707063,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979941950554195,
      "loss": 5.4654,
      "mean_token_accuracy": 0.17086824923753738,
      "num_tokens": 11158549.0,
      "step": 6425
    },
    {
      "entropy": 5.647154426574707,
      "epoch": 0.5002917720287882,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979905008499399,
      "loss": 5.4042,
      "mean_token_accuracy": 0.1790368750691414,
      "num_tokens": 11167131.0,
      "step": 6430
    },
    {
      "entropy": 5.637821531295776,
      "epoch": 0.5006808014005057,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979868032609344,
      "loss": 5.4883,
      "mean_token_accuracy": 0.16543890535831451,
      "num_tokens": 11175766.0,
      "step": 6435
    },
    {
      "entropy": 5.635375261306763,
      "epoch": 0.5010698307722233,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979831022884591,
      "loss": 5.451,
      "mean_token_accuracy": 0.17667505741119385,
      "num_tokens": 11184799.0,
      "step": 6440
    },
    {
      "entropy": 5.610274791717529,
      "epoch": 0.5014588601439409,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979793979325701,
      "loss": 5.3767,
      "mean_token_accuracy": 0.16970710307359696,
      "num_tokens": 11193083.0,
      "step": 6445
    },
    {
      "entropy": 5.606352710723877,
      "epoch": 0.5018478895156584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979756901933236,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17012580037117003,
      "num_tokens": 11201753.0,
      "step": 6450
    },
    {
      "entropy": 5.6947142601013185,
      "epoch": 0.502236918887376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000497971979070776,
      "loss": 5.4968,
      "mean_token_accuracy": 0.1657540574669838,
      "num_tokens": 11209965.0,
      "step": 6455
    },
    {
      "entropy": 5.6152198791503904,
      "epoch": 0.5026259482590936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979682645649834,
      "loss": 5.3652,
      "mean_token_accuracy": 0.17716653347015382,
      "num_tokens": 11217961.0,
      "step": 6460
    },
    {
      "entropy": 5.634564304351807,
      "epoch": 0.5030149776308112,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004979645466760025,
      "loss": 5.5001,
      "mean_token_accuracy": 0.16121395379304887,
      "num_tokens": 11225803.0,
      "step": 6465
    },
    {
      "entropy": 5.637865304946899,
      "epoch": 0.5034040070025287,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979608254038892,
      "loss": 5.3456,
      "mean_token_accuracy": 0.17098010927438737,
      "num_tokens": 11234584.0,
      "step": 6470
    },
    {
      "entropy": 5.71119327545166,
      "epoch": 0.5037930363742462,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979571007487003,
      "loss": 5.536,
      "mean_token_accuracy": 0.16245963275432587,
      "num_tokens": 11243340.0,
      "step": 6475
    },
    {
      "entropy": 5.619586372375489,
      "epoch": 0.5041820657459638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979533727104924,
      "loss": 5.3997,
      "mean_token_accuracy": 0.16557498127222062,
      "num_tokens": 11252383.0,
      "step": 6480
    },
    {
      "entropy": 5.646946907043457,
      "epoch": 0.5045710951176814,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497949641289322,
      "loss": 5.4775,
      "mean_token_accuracy": 0.16912005543708802,
      "num_tokens": 11261419.0,
      "step": 6485
    },
    {
      "entropy": 5.654446411132812,
      "epoch": 0.504960124489399,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979459064852456,
      "loss": 5.4496,
      "mean_token_accuracy": 0.16540549248456954,
      "num_tokens": 11270743.0,
      "step": 6490
    },
    {
      "entropy": 5.638967895507813,
      "epoch": 0.5053491538611166,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979421682983197,
      "loss": 5.3424,
      "mean_token_accuracy": 0.17329360842704772,
      "num_tokens": 11279666.0,
      "step": 6495
    },
    {
      "entropy": 5.669098615646362,
      "epoch": 0.505738183232834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979384267286015,
      "loss": 5.5363,
      "mean_token_accuracy": 0.16254350841045379,
      "num_tokens": 11288710.0,
      "step": 6500
    },
    {
      "entropy": 5.661252021789551,
      "epoch": 0.5061272126045516,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979346817761475,
      "loss": 5.4092,
      "mean_token_accuracy": 0.16631939113140107,
      "num_tokens": 11297665.0,
      "step": 6505
    },
    {
      "entropy": 5.697106981277466,
      "epoch": 0.5065162419762692,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979309334410144,
      "loss": 5.5657,
      "mean_token_accuracy": 0.15991870909929276,
      "num_tokens": 11306751.0,
      "step": 6510
    },
    {
      "entropy": 5.708811950683594,
      "epoch": 0.5069052713479868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979271817232594,
      "loss": 5.5537,
      "mean_token_accuracy": 0.15796037316322326,
      "num_tokens": 11315659.0,
      "step": 6515
    },
    {
      "entropy": 5.689514636993408,
      "epoch": 0.5072943007197044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979234266229391,
      "loss": 5.4448,
      "mean_token_accuracy": 0.1685910552740097,
      "num_tokens": 11323362.0,
      "step": 6520
    },
    {
      "entropy": 5.656657600402832,
      "epoch": 0.5076833300914219,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979196681401106,
      "loss": 5.4477,
      "mean_token_accuracy": 0.1603214606642723,
      "num_tokens": 11332343.0,
      "step": 6525
    },
    {
      "entropy": 5.6222803592681885,
      "epoch": 0.5080723594631394,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497915906274831,
      "loss": 5.3762,
      "mean_token_accuracy": 0.16662061214447021,
      "num_tokens": 11340816.0,
      "step": 6530
    },
    {
      "entropy": 5.590240573883056,
      "epoch": 0.508461388834857,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979121410271574,
      "loss": 5.3876,
      "mean_token_accuracy": 0.16722644865512848,
      "num_tokens": 11349728.0,
      "step": 6535
    },
    {
      "entropy": 5.669598007202149,
      "epoch": 0.5088504182065746,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979083723971468,
      "loss": 5.4995,
      "mean_token_accuracy": 0.16796678602695464,
      "num_tokens": 11358477.0,
      "step": 6540
    },
    {
      "entropy": 5.5297510623931885,
      "epoch": 0.5092394475782922,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979046003848564,
      "loss": 5.3225,
      "mean_token_accuracy": 0.17740241587162017,
      "num_tokens": 11367183.0,
      "step": 6545
    },
    {
      "entropy": 5.656442737579345,
      "epoch": 0.5096284769500097,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979008249903435,
      "loss": 5.5264,
      "mean_token_accuracy": 0.15788258463144303,
      "num_tokens": 11376703.0,
      "step": 6550
    },
    {
      "entropy": 5.7535947322845455,
      "epoch": 0.5100175063217273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978970462136655,
      "loss": 5.5325,
      "mean_token_accuracy": 0.16580595076084137,
      "num_tokens": 11385502.0,
      "step": 6555
    },
    {
      "entropy": 5.637090444564819,
      "epoch": 0.5104065356934449,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978932640548794,
      "loss": 5.3987,
      "mean_token_accuracy": 0.17354696840047837,
      "num_tokens": 11393947.0,
      "step": 6560
    },
    {
      "entropy": 5.577970886230469,
      "epoch": 0.5107955650651624,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978894785140429,
      "loss": 5.3901,
      "mean_token_accuracy": 0.17127427607774734,
      "num_tokens": 11403077.0,
      "step": 6565
    },
    {
      "entropy": 5.572803497314453,
      "epoch": 0.51118459443688,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978856895912132,
      "loss": 5.3867,
      "mean_token_accuracy": 0.1690242201089859,
      "num_tokens": 11411863.0,
      "step": 6570
    },
    {
      "entropy": 5.644185352325439,
      "epoch": 0.5115736238085975,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978818972864481,
      "loss": 5.3522,
      "mean_token_accuracy": 0.1726667046546936,
      "num_tokens": 11420030.0,
      "step": 6575
    },
    {
      "entropy": 5.679980611801147,
      "epoch": 0.5119626531803151,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978781015998048,
      "loss": 5.429,
      "mean_token_accuracy": 0.16585396677255632,
      "num_tokens": 11428534.0,
      "step": 6580
    },
    {
      "entropy": 5.6785746097564695,
      "epoch": 0.5123516825520327,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978743025313413,
      "loss": 5.4153,
      "mean_token_accuracy": 0.17318394929170608,
      "num_tokens": 11436641.0,
      "step": 6585
    },
    {
      "entropy": 5.611976623535156,
      "epoch": 0.5127407119237503,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978705000811148,
      "loss": 5.4042,
      "mean_token_accuracy": 0.1705085426568985,
      "num_tokens": 11444595.0,
      "step": 6590
    },
    {
      "entropy": 5.578699254989624,
      "epoch": 0.5131297412954678,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978666942491832,
      "loss": 5.2582,
      "mean_token_accuracy": 0.1794433295726776,
      "num_tokens": 11453257.0,
      "step": 6595
    },
    {
      "entropy": 5.621994495391846,
      "epoch": 0.5135187706671853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978628850356043,
      "loss": 5.4233,
      "mean_token_accuracy": 0.17493958324193953,
      "num_tokens": 11461762.0,
      "step": 6600
    },
    {
      "entropy": 5.625024843215942,
      "epoch": 0.5139078000389029,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978590724404358,
      "loss": 5.3386,
      "mean_token_accuracy": 0.17288504540920258,
      "num_tokens": 11470602.0,
      "step": 6605
    },
    {
      "entropy": 5.679289627075195,
      "epoch": 0.5142968294106205,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978552564637356,
      "loss": 5.4435,
      "mean_token_accuracy": 0.16734055280685425,
      "num_tokens": 11479826.0,
      "step": 6610
    },
    {
      "entropy": 5.6650903701782225,
      "epoch": 0.5146858587823381,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978514371055616,
      "loss": 5.4934,
      "mean_token_accuracy": 0.16676056981086732,
      "num_tokens": 11488698.0,
      "step": 6615
    },
    {
      "entropy": 5.700325441360474,
      "epoch": 0.5150748881540557,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978476143659718,
      "loss": 5.4767,
      "mean_token_accuracy": 0.1701172336935997,
      "num_tokens": 11497434.0,
      "step": 6620
    },
    {
      "entropy": 5.6457771301269535,
      "epoch": 0.5154639175257731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978437882450241,
      "loss": 5.4921,
      "mean_token_accuracy": 0.16389915943145753,
      "num_tokens": 11506173.0,
      "step": 6625
    },
    {
      "entropy": 5.666603851318359,
      "epoch": 0.5158529468974907,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978399587427766,
      "loss": 5.4015,
      "mean_token_accuracy": 0.17104636281728744,
      "num_tokens": 11514824.0,
      "step": 6630
    },
    {
      "entropy": 5.634468507766724,
      "epoch": 0.5162419762692083,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978361258592874,
      "loss": 5.459,
      "mean_token_accuracy": 0.15855541080236435,
      "num_tokens": 11523634.0,
      "step": 6635
    },
    {
      "entropy": 5.744538164138794,
      "epoch": 0.5166310056409259,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978322895946147,
      "loss": 5.5689,
      "mean_token_accuracy": 0.16215945929288864,
      "num_tokens": 11531916.0,
      "step": 6640
    },
    {
      "entropy": 5.667070627212524,
      "epoch": 0.5170200350126435,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004978284499488167,
      "loss": 5.3887,
      "mean_token_accuracy": 0.16777578592300416,
      "num_tokens": 11540000.0,
      "step": 6645
    },
    {
      "entropy": 5.720247840881347,
      "epoch": 0.5174090643843611,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978246069219516,
      "loss": 5.4711,
      "mean_token_accuracy": 0.16854884028434752,
      "num_tokens": 11548760.0,
      "step": 6650
    },
    {
      "entropy": 5.600211334228516,
      "epoch": 0.5177980937560785,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004978207605140777,
      "loss": 5.3013,
      "mean_token_accuracy": 0.17687248587608337,
      "num_tokens": 11557991.0,
      "step": 6655
    },
    {
      "entropy": 5.582965087890625,
      "epoch": 0.5181871231277961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978169107252534,
      "loss": 5.375,
      "mean_token_accuracy": 0.17161434441804885,
      "num_tokens": 11566638.0,
      "step": 6660
    },
    {
      "entropy": 5.529303884506225,
      "epoch": 0.5185761524995137,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978130575555373,
      "loss": 5.3102,
      "mean_token_accuracy": 0.1714066445827484,
      "num_tokens": 11575106.0,
      "step": 6665
    },
    {
      "entropy": 5.571141386032105,
      "epoch": 0.5189651818712313,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978092010049877,
      "loss": 5.4538,
      "mean_token_accuracy": 0.1577305719256401,
      "num_tokens": 11583971.0,
      "step": 6670
    },
    {
      "entropy": 5.640804672241211,
      "epoch": 0.5193542112429489,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497805341073663,
      "loss": 5.4474,
      "mean_token_accuracy": 0.16622724384069443,
      "num_tokens": 11592936.0,
      "step": 6675
    },
    {
      "entropy": 5.64942626953125,
      "epoch": 0.5197432406146664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978014777616219,
      "loss": 5.3465,
      "mean_token_accuracy": 0.1758352756500244,
      "num_tokens": 11602055.0,
      "step": 6680
    },
    {
      "entropy": 5.631604099273682,
      "epoch": 0.520132269986384,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977976110689229,
      "loss": 5.3533,
      "mean_token_accuracy": 0.17612868696451187,
      "num_tokens": 11610211.0,
      "step": 6685
    },
    {
      "entropy": 5.666025066375733,
      "epoch": 0.5205212993581015,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497793740995625,
      "loss": 5.4378,
      "mean_token_accuracy": 0.1702517494559288,
      "num_tokens": 11619096.0,
      "step": 6690
    },
    {
      "entropy": 5.6196661472320555,
      "epoch": 0.5209103287298191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977898675417866,
      "loss": 5.5078,
      "mean_token_accuracy": 0.17089473754167556,
      "num_tokens": 11628142.0,
      "step": 6695
    },
    {
      "entropy": 5.675712394714355,
      "epoch": 0.5212993581015367,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977859907074664,
      "loss": 5.3536,
      "mean_token_accuracy": 0.1730503797531128,
      "num_tokens": 11635961.0,
      "step": 6700
    },
    {
      "entropy": 5.674269342422486,
      "epoch": 0.5216883874732542,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977821104927236,
      "loss": 5.4708,
      "mean_token_accuracy": 0.1689344897866249,
      "num_tokens": 11644778.0,
      "step": 6705
    },
    {
      "entropy": 5.628959465026855,
      "epoch": 0.5220774168449718,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977782268976167,
      "loss": 5.4639,
      "mean_token_accuracy": 0.16584417819976807,
      "num_tokens": 11654043.0,
      "step": 6710
    },
    {
      "entropy": 5.666242885589599,
      "epoch": 0.5224664462166894,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977743399222048,
      "loss": 5.4486,
      "mean_token_accuracy": 0.16625992506742476,
      "num_tokens": 11662794.0,
      "step": 6715
    },
    {
      "entropy": 5.605064916610718,
      "epoch": 0.522855475588407,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977704495665471,
      "loss": 5.4409,
      "mean_token_accuracy": 0.17458575814962388,
      "num_tokens": 11670624.0,
      "step": 6720
    },
    {
      "entropy": 5.560970878601074,
      "epoch": 0.5232445049601245,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977665558307023,
      "loss": 5.3436,
      "mean_token_accuracy": 0.1714425042271614,
      "num_tokens": 11679366.0,
      "step": 6725
    },
    {
      "entropy": 5.57506365776062,
      "epoch": 0.523633534331842,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977626587147295,
      "loss": 5.2532,
      "mean_token_accuracy": 0.17893183827400208,
      "num_tokens": 11688137.0,
      "step": 6730
    },
    {
      "entropy": 5.5754714012146,
      "epoch": 0.5240225637035596,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497758758218688,
      "loss": 5.3879,
      "mean_token_accuracy": 0.1726705551147461,
      "num_tokens": 11697236.0,
      "step": 6735
    },
    {
      "entropy": 5.620447826385498,
      "epoch": 0.5244115930752772,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977548543426368,
      "loss": 5.3889,
      "mean_token_accuracy": 0.1749192327260971,
      "num_tokens": 11706513.0,
      "step": 6740
    },
    {
      "entropy": 5.649651384353637,
      "epoch": 0.5248006224469948,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977509470866353,
      "loss": 5.3506,
      "mean_token_accuracy": 0.17255444675683976,
      "num_tokens": 11715239.0,
      "step": 6745
    },
    {
      "entropy": 5.605628538131714,
      "epoch": 0.5251896518187124,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977470364507427,
      "loss": 5.3942,
      "mean_token_accuracy": 0.17865999937057495,
      "num_tokens": 11724086.0,
      "step": 6750
    },
    {
      "entropy": 5.576850509643554,
      "epoch": 0.5255786811904298,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977431224350184,
      "loss": 5.4121,
      "mean_token_accuracy": 0.17535001784563065,
      "num_tokens": 11731975.0,
      "step": 6755
    },
    {
      "entropy": 5.5735608577728275,
      "epoch": 0.5259677105621474,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977392050395217,
      "loss": 5.4015,
      "mean_token_accuracy": 0.1689216673374176,
      "num_tokens": 11741172.0,
      "step": 6760
    },
    {
      "entropy": 5.618299150466919,
      "epoch": 0.526356739933865,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977352842643121,
      "loss": 5.3595,
      "mean_token_accuracy": 0.17527500689029693,
      "num_tokens": 11750609.0,
      "step": 6765
    },
    {
      "entropy": 5.5818558692932125,
      "epoch": 0.5267457693055826,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497731360109449,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1736464872956276,
      "num_tokens": 11758901.0,
      "step": 6770
    },
    {
      "entropy": 5.52725043296814,
      "epoch": 0.5271347986773002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977274325749922,
      "loss": 5.3077,
      "mean_token_accuracy": 0.17586426734924315,
      "num_tokens": 11767202.0,
      "step": 6775
    },
    {
      "entropy": 5.585738468170166,
      "epoch": 0.5275238280490177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977235016610011,
      "loss": 5.4002,
      "mean_token_accuracy": 0.17273519486188887,
      "num_tokens": 11776164.0,
      "step": 6780
    },
    {
      "entropy": 5.578510379791259,
      "epoch": 0.5279128574207352,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004977195673675353,
      "loss": 5.3631,
      "mean_token_accuracy": 0.17961377501487732,
      "num_tokens": 11784548.0,
      "step": 6785
    },
    {
      "entropy": 5.582890176773072,
      "epoch": 0.5283018867924528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004977156296946545,
      "loss": 5.4242,
      "mean_token_accuracy": 0.17239639312028884,
      "num_tokens": 11793249.0,
      "step": 6790
    },
    {
      "entropy": 5.684317922592163,
      "epoch": 0.5286909161641704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977116886424187,
      "loss": 5.475,
      "mean_token_accuracy": 0.15826374739408494,
      "num_tokens": 11801512.0,
      "step": 6795
    },
    {
      "entropy": 5.612030124664306,
      "epoch": 0.529079945535888,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977077442108873,
      "loss": 5.3535,
      "mean_token_accuracy": 0.174309204518795,
      "num_tokens": 11810248.0,
      "step": 6800
    },
    {
      "entropy": 5.599524450302124,
      "epoch": 0.5294689749076055,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977037964001204,
      "loss": 5.3448,
      "mean_token_accuracy": 0.171326445043087,
      "num_tokens": 11819190.0,
      "step": 6805
    },
    {
      "entropy": 5.583758401870727,
      "epoch": 0.5298580042793231,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497699845210178,
      "loss": 5.4194,
      "mean_token_accuracy": 0.16252280473709108,
      "num_tokens": 11827585.0,
      "step": 6810
    },
    {
      "entropy": 5.6619100093841555,
      "epoch": 0.5302470336510406,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004976958906411197,
      "loss": 5.4057,
      "mean_token_accuracy": 0.17098232507705688,
      "num_tokens": 11836256.0,
      "step": 6815
    },
    {
      "entropy": 5.672984313964844,
      "epoch": 0.5306360630227582,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976919326930059,
      "loss": 5.5252,
      "mean_token_accuracy": 0.17031456232070924,
      "num_tokens": 11845022.0,
      "step": 6820
    },
    {
      "entropy": 5.6437376022338865,
      "epoch": 0.5310250923944758,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976879713658964,
      "loss": 5.3832,
      "mean_token_accuracy": 0.1698227718472481,
      "num_tokens": 11853289.0,
      "step": 6825
    },
    {
      "entropy": 5.568662071228028,
      "epoch": 0.5314141217661933,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976840066598514,
      "loss": 5.3322,
      "mean_token_accuracy": 0.17991804629564284,
      "num_tokens": 11861589.0,
      "step": 6830
    },
    {
      "entropy": 5.664581918716431,
      "epoch": 0.5318031511379109,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976800385749309,
      "loss": 5.4406,
      "mean_token_accuracy": 0.17165744453668594,
      "num_tokens": 11870313.0,
      "step": 6835
    },
    {
      "entropy": 5.7209772109985355,
      "epoch": 0.5321921805096285,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976760671111954,
      "loss": 5.4333,
      "mean_token_accuracy": 0.16864327043294908,
      "num_tokens": 11879558.0,
      "step": 6840
    },
    {
      "entropy": 5.6440380096435545,
      "epoch": 0.532581209881346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497672092268705,
      "loss": 5.4412,
      "mean_token_accuracy": 0.1721239373087883,
      "num_tokens": 11887951.0,
      "step": 6845
    },
    {
      "entropy": 5.696996545791626,
      "epoch": 0.5329702392530636,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049766811404752,
      "loss": 5.638,
      "mean_token_accuracy": 0.15808830112218858,
      "num_tokens": 11896900.0,
      "step": 6850
    },
    {
      "entropy": 5.677005386352539,
      "epoch": 0.5333592686247812,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004976641324477008,
      "loss": 5.469,
      "mean_token_accuracy": 0.1657126635313034,
      "num_tokens": 11905681.0,
      "step": 6855
    },
    {
      "entropy": 5.612731027603149,
      "epoch": 0.5337482979964987,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976601474693077,
      "loss": 5.4003,
      "mean_token_accuracy": 0.17090844810009004,
      "num_tokens": 11914662.0,
      "step": 6860
    },
    {
      "entropy": 5.567288160324097,
      "epoch": 0.5341373273682163,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004976561591124014,
      "loss": 5.4053,
      "mean_token_accuracy": 0.17699283957481385,
      "num_tokens": 11923298.0,
      "step": 6865
    },
    {
      "entropy": 5.639786863327027,
      "epoch": 0.5345263567399339,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976521673770421,
      "loss": 5.3198,
      "mean_token_accuracy": 0.17048582732677459,
      "num_tokens": 11932206.0,
      "step": 6870
    },
    {
      "entropy": 5.580082273483276,
      "epoch": 0.5349153861116515,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976481722632907,
      "loss": 5.3763,
      "mean_token_accuracy": 0.176534940302372,
      "num_tokens": 11940786.0,
      "step": 6875
    },
    {
      "entropy": 5.612260055541992,
      "epoch": 0.535304415483369,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976441737712076,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1711177110671997,
      "num_tokens": 11949619.0,
      "step": 6880
    },
    {
      "entropy": 5.603137588500976,
      "epoch": 0.5356934448550865,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004976401719008536,
      "loss": 5.3338,
      "mean_token_accuracy": 0.17791433930397033,
      "num_tokens": 11957696.0,
      "step": 6885
    },
    {
      "entropy": 5.576803970336914,
      "epoch": 0.5360824742268041,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004976361666522893,
      "loss": 5.3404,
      "mean_token_accuracy": 0.1796420320868492,
      "num_tokens": 11965622.0,
      "step": 6890
    },
    {
      "entropy": 5.563084983825684,
      "epoch": 0.5364715035985217,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976321580255755,
      "loss": 5.4238,
      "mean_token_accuracy": 0.17168654203414918,
      "num_tokens": 11973969.0,
      "step": 6895
    },
    {
      "entropy": 5.705384826660156,
      "epoch": 0.5368605329702393,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976281460207731,
      "loss": 5.41,
      "mean_token_accuracy": 0.1748523712158203,
      "num_tokens": 11982469.0,
      "step": 6900
    },
    {
      "entropy": 5.688539218902588,
      "epoch": 0.5372495623419569,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497624130637943,
      "loss": 5.5045,
      "mean_token_accuracy": 0.16434845477342605,
      "num_tokens": 11990656.0,
      "step": 6905
    },
    {
      "entropy": 5.645046281814575,
      "epoch": 0.5376385917136743,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497620111877146,
      "loss": 5.4958,
      "mean_token_accuracy": 0.16131058484315872,
      "num_tokens": 11999394.0,
      "step": 6910
    },
    {
      "entropy": 5.685613632202148,
      "epoch": 0.5380276210853919,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004976160897384431,
      "loss": 5.4609,
      "mean_token_accuracy": 0.17145871371030807,
      "num_tokens": 12007692.0,
      "step": 6915
    },
    {
      "entropy": 5.661275339126587,
      "epoch": 0.5384166504571095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976120642218955,
      "loss": 5.4336,
      "mean_token_accuracy": 0.16824211180210114,
      "num_tokens": 12016038.0,
      "step": 6920
    },
    {
      "entropy": 5.650868272781372,
      "epoch": 0.5388056798288271,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497608035327564,
      "loss": 5.5396,
      "mean_token_accuracy": 0.16747572720050813,
      "num_tokens": 12024504.0,
      "step": 6925
    },
    {
      "entropy": 5.608466005325317,
      "epoch": 0.5391947092005447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049760400305551,
      "loss": 5.4369,
      "mean_token_accuracy": 0.17732071429491042,
      "num_tokens": 12032935.0,
      "step": 6930
    },
    {
      "entropy": 5.65435676574707,
      "epoch": 0.5395837385722622,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975999674057944,
      "loss": 5.4592,
      "mean_token_accuracy": 0.16456467062234878,
      "num_tokens": 12041907.0,
      "step": 6935
    },
    {
      "entropy": 5.607891321182251,
      "epoch": 0.5399727679439797,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975959283784787,
      "loss": 5.2978,
      "mean_token_accuracy": 0.17626878917217254,
      "num_tokens": 12050528.0,
      "step": 6940
    },
    {
      "entropy": 5.570985174179077,
      "epoch": 0.5403617973156973,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975918859736241,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1774978369474411,
      "num_tokens": 12059221.0,
      "step": 6945
    },
    {
      "entropy": 5.630969619750976,
      "epoch": 0.5407508266874149,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975878401912919,
      "loss": 5.5186,
      "mean_token_accuracy": 0.16120119094848634,
      "num_tokens": 12068131.0,
      "step": 6950
    },
    {
      "entropy": 5.698724842071533,
      "epoch": 0.5411398560591325,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975837910315436,
      "loss": 5.3707,
      "mean_token_accuracy": 0.18022255450487137,
      "num_tokens": 12076521.0,
      "step": 6955
    },
    {
      "entropy": 5.683851289749145,
      "epoch": 0.54152888543085,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975797384944405,
      "loss": 5.5697,
      "mean_token_accuracy": 0.16650937199592591,
      "num_tokens": 12085306.0,
      "step": 6960
    },
    {
      "entropy": 5.593390607833863,
      "epoch": 0.5419179148025676,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497575682580044,
      "loss": 5.3415,
      "mean_token_accuracy": 0.16895500123500823,
      "num_tokens": 12094409.0,
      "step": 6965
    },
    {
      "entropy": 5.617079973220825,
      "epoch": 0.5423069441742852,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497571623288416,
      "loss": 5.4016,
      "mean_token_accuracy": 0.16630526036024093,
      "num_tokens": 12103385.0,
      "step": 6970
    },
    {
      "entropy": 5.682160234451294,
      "epoch": 0.5426959735460027,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004975675606196177,
      "loss": 5.5081,
      "mean_token_accuracy": 0.1710004076361656,
      "num_tokens": 12112452.0,
      "step": 6975
    },
    {
      "entropy": 5.735922050476074,
      "epoch": 0.5430850029177203,
      "grad_norm": 1.0,
      "learning_rate": 0.000497563494573711,
      "loss": 5.501,
      "mean_token_accuracy": 0.16131800562143325,
      "num_tokens": 12120918.0,
      "step": 6980
    },
    {
      "entropy": 5.659406328201294,
      "epoch": 0.5434740322894378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975594251507575,
      "loss": 5.4709,
      "mean_token_accuracy": 0.16228864043951036,
      "num_tokens": 12129797.0,
      "step": 6985
    },
    {
      "entropy": 5.619140863418579,
      "epoch": 0.5438630616611554,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497555352350819,
      "loss": 5.4487,
      "mean_token_accuracy": 0.16991385221481323,
      "num_tokens": 12139245.0,
      "step": 6990
    },
    {
      "entropy": 5.685261535644531,
      "epoch": 0.544252091032873,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975512761739572,
      "loss": 5.6102,
      "mean_token_accuracy": 0.16088290214538575,
      "num_tokens": 12147409.0,
      "step": 6995
    },
    {
      "entropy": 5.724803781509399,
      "epoch": 0.5446411204045906,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497547196620234,
      "loss": 5.424,
      "mean_token_accuracy": 0.17047625184059143,
      "num_tokens": 12156121.0,
      "step": 7000
    },
    {
      "entropy": 5.586559152603149,
      "epoch": 0.5450301497763081,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975431136897114,
      "loss": 5.3186,
      "mean_token_accuracy": 0.17673922777175904,
      "num_tokens": 12165307.0,
      "step": 7005
    },
    {
      "entropy": 5.633096408843994,
      "epoch": 0.5454191791480256,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975390273824512,
      "loss": 5.3832,
      "mean_token_accuracy": 0.1656043067574501,
      "num_tokens": 12173899.0,
      "step": 7010
    },
    {
      "entropy": 5.560071325302124,
      "epoch": 0.5458082085197432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975349376985155,
      "loss": 5.3972,
      "mean_token_accuracy": 0.16830164492130278,
      "num_tokens": 12182898.0,
      "step": 7015
    },
    {
      "entropy": 5.542508268356324,
      "epoch": 0.5461972378914608,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975308446379663,
      "loss": 5.275,
      "mean_token_accuracy": 0.1712779715657234,
      "num_tokens": 12191500.0,
      "step": 7020
    },
    {
      "entropy": 5.579271554946899,
      "epoch": 0.5465862672631784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975267482008657,
      "loss": 5.4364,
      "mean_token_accuracy": 0.17276248931884766,
      "num_tokens": 12199420.0,
      "step": 7025
    },
    {
      "entropy": 5.6312158584594725,
      "epoch": 0.546975296634896,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497522648387276,
      "loss": 5.3523,
      "mean_token_accuracy": 0.17430415749549866,
      "num_tokens": 12207676.0,
      "step": 7030
    },
    {
      "entropy": 5.647388362884522,
      "epoch": 0.5473643260066134,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975185451972592,
      "loss": 5.4124,
      "mean_token_accuracy": 0.1714552164077759,
      "num_tokens": 12216248.0,
      "step": 7035
    },
    {
      "entropy": 5.624428606033325,
      "epoch": 0.547753355378331,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975144386308776,
      "loss": 5.4368,
      "mean_token_accuracy": 0.17158026099205018,
      "num_tokens": 12224588.0,
      "step": 7040
    },
    {
      "entropy": 5.657812929153442,
      "epoch": 0.5481423847500486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975103286881936,
      "loss": 5.4578,
      "mean_token_accuracy": 0.1659623846411705,
      "num_tokens": 12234093.0,
      "step": 7045
    },
    {
      "entropy": 5.721933794021607,
      "epoch": 0.5485314141217662,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975062153692696,
      "loss": 5.3901,
      "mean_token_accuracy": 0.18022023141384125,
      "num_tokens": 12242164.0,
      "step": 7050
    },
    {
      "entropy": 5.649906730651855,
      "epoch": 0.5489204434934838,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004975020986741678,
      "loss": 5.4898,
      "mean_token_accuracy": 0.16861753314733505,
      "num_tokens": 12251204.0,
      "step": 7055
    },
    {
      "entropy": 5.6687757015228275,
      "epoch": 0.5493094728652014,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004974979786029509,
      "loss": 5.4202,
      "mean_token_accuracy": 0.1690824121236801,
      "num_tokens": 12259809.0,
      "step": 7060
    },
    {
      "entropy": 5.624884033203125,
      "epoch": 0.5496985022369189,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974938551556814,
      "loss": 5.4069,
      "mean_token_accuracy": 0.17586851567029954,
      "num_tokens": 12268668.0,
      "step": 7065
    },
    {
      "entropy": 5.613442516326904,
      "epoch": 0.5500875316086364,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974897283324217,
      "loss": 5.3819,
      "mean_token_accuracy": 0.17798949629068375,
      "num_tokens": 12276823.0,
      "step": 7070
    },
    {
      "entropy": 5.485236740112304,
      "epoch": 0.550476560980354,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974855981332343,
      "loss": 5.3098,
      "mean_token_accuracy": 0.1711676076054573,
      "num_tokens": 12285950.0,
      "step": 7075
    },
    {
      "entropy": 5.597700595855713,
      "epoch": 0.5508655903520716,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974814645581823,
      "loss": 5.4277,
      "mean_token_accuracy": 0.16414272636175156,
      "num_tokens": 12295265.0,
      "step": 7080
    },
    {
      "entropy": 5.662305927276611,
      "epoch": 0.5512546197237892,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004974773276073282,
      "loss": 5.4386,
      "mean_token_accuracy": 0.1707385465502739,
      "num_tokens": 12303649.0,
      "step": 7085
    },
    {
      "entropy": 5.59878978729248,
      "epoch": 0.5516436490955067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004974731872807347,
      "loss": 5.4989,
      "mean_token_accuracy": 0.16129840314388275,
      "num_tokens": 12313303.0,
      "step": 7090
    },
    {
      "entropy": 5.6164943218231205,
      "epoch": 0.5520326784672243,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974690435784647,
      "loss": 5.4422,
      "mean_token_accuracy": 0.1737209215760231,
      "num_tokens": 12321692.0,
      "step": 7095
    },
    {
      "entropy": 5.521945524215698,
      "epoch": 0.5524217078389418,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974648965005811,
      "loss": 5.2479,
      "mean_token_accuracy": 0.17637658715248108,
      "num_tokens": 12329908.0,
      "step": 7100
    },
    {
      "entropy": 5.562238073348999,
      "epoch": 0.5528107372106594,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974607460471466,
      "loss": 5.4201,
      "mean_token_accuracy": 0.1720580741763115,
      "num_tokens": 12339270.0,
      "step": 7105
    },
    {
      "entropy": 5.589144849777222,
      "epoch": 0.553199766582377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974565922182246,
      "loss": 5.3241,
      "mean_token_accuracy": 0.17368963211774827,
      "num_tokens": 12347586.0,
      "step": 7110
    },
    {
      "entropy": 5.549763822555542,
      "epoch": 0.5535887959540945,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974524350138777,
      "loss": 5.284,
      "mean_token_accuracy": 0.170064477622509,
      "num_tokens": 12356291.0,
      "step": 7115
    },
    {
      "entropy": 5.640159797668457,
      "epoch": 0.5539778253258121,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974482744341693,
      "loss": 5.4606,
      "mean_token_accuracy": 0.1634287178516388,
      "num_tokens": 12364948.0,
      "step": 7120
    },
    {
      "entropy": 5.6569324970245365,
      "epoch": 0.5543668546975297,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974441104791624,
      "loss": 5.407,
      "mean_token_accuracy": 0.17270324677228927,
      "num_tokens": 12373497.0,
      "step": 7125
    },
    {
      "entropy": 5.6427390575408936,
      "epoch": 0.5547558840692473,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004974399431489201,
      "loss": 5.3587,
      "mean_token_accuracy": 0.17324348986148835,
      "num_tokens": 12382198.0,
      "step": 7130
    },
    {
      "entropy": 5.5707135677337645,
      "epoch": 0.5551449134409648,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004974357724435057,
      "loss": 5.3478,
      "mean_token_accuracy": 0.17520179748535156,
      "num_tokens": 12391122.0,
      "step": 7135
    },
    {
      "entropy": 5.504246950149536,
      "epoch": 0.5555339428126823,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974315983629825,
      "loss": 5.291,
      "mean_token_accuracy": 0.18072783648967744,
      "num_tokens": 12399338.0,
      "step": 7140
    },
    {
      "entropy": 5.615718746185303,
      "epoch": 0.5559229721843999,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974274209074139,
      "loss": 5.3896,
      "mean_token_accuracy": 0.16674543172121048,
      "num_tokens": 12408561.0,
      "step": 7145
    },
    {
      "entropy": 5.643730163574219,
      "epoch": 0.5563120015561175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974232400768632,
      "loss": 5.2761,
      "mean_token_accuracy": 0.1755574345588684,
      "num_tokens": 12417727.0,
      "step": 7150
    },
    {
      "entropy": 5.601311254501343,
      "epoch": 0.5567010309278351,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974190558713939,
      "loss": 5.3009,
      "mean_token_accuracy": 0.17797841727733613,
      "num_tokens": 12426446.0,
      "step": 7155
    },
    {
      "entropy": 5.574703121185303,
      "epoch": 0.5570900602995527,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974148682910694,
      "loss": 5.3702,
      "mean_token_accuracy": 0.169679394364357,
      "num_tokens": 12435124.0,
      "step": 7160
    },
    {
      "entropy": 5.632902956008911,
      "epoch": 0.5574790896712701,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004974106773359533,
      "loss": 5.4616,
      "mean_token_accuracy": 0.16738623529672622,
      "num_tokens": 12444073.0,
      "step": 7165
    },
    {
      "entropy": 5.694697999954224,
      "epoch": 0.5578681190429877,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004974064830061091,
      "loss": 5.4502,
      "mean_token_accuracy": 0.17598361372947693,
      "num_tokens": 12452419.0,
      "step": 7170
    },
    {
      "entropy": 5.608331632614136,
      "epoch": 0.5582571484147053,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974022853016006,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17239612936973572,
      "num_tokens": 12461488.0,
      "step": 7175
    },
    {
      "entropy": 5.6137001514434814,
      "epoch": 0.5586461777864229,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973980842224913,
      "loss": 5.3359,
      "mean_token_accuracy": 0.16857871264219285,
      "num_tokens": 12470344.0,
      "step": 7180
    },
    {
      "entropy": 5.608567142486573,
      "epoch": 0.5590352071581405,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973938797688452,
      "loss": 5.347,
      "mean_token_accuracy": 0.17229609936475754,
      "num_tokens": 12479189.0,
      "step": 7185
    },
    {
      "entropy": 5.599718999862671,
      "epoch": 0.559424236529858,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004973896719407259,
      "loss": 5.3849,
      "mean_token_accuracy": 0.17104582041501998,
      "num_tokens": 12488199.0,
      "step": 7190
    },
    {
      "entropy": 5.544893550872803,
      "epoch": 0.5598132659015755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004973854607381972,
      "loss": 5.2665,
      "mean_token_accuracy": 0.18018234372138978,
      "num_tokens": 12496704.0,
      "step": 7195
    },
    {
      "entropy": 5.496060514450074,
      "epoch": 0.5602022952732931,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004973812461613232,
      "loss": 5.2633,
      "mean_token_accuracy": 0.18417042195796968,
      "num_tokens": 12504724.0,
      "step": 7200
    },
    {
      "entropy": 5.59717960357666,
      "epoch": 0.5605913246450107,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973770282101677,
      "loss": 5.4681,
      "mean_token_accuracy": 0.1670819789171219,
      "num_tokens": 12514614.0,
      "step": 7205
    },
    {
      "entropy": 5.747423458099365,
      "epoch": 0.5609803540167283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004973728068847947,
      "loss": 5.4762,
      "mean_token_accuracy": 0.1688718467950821,
      "num_tokens": 12522744.0,
      "step": 7210
    },
    {
      "entropy": 5.6050097942352295,
      "epoch": 0.5613693833884458,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004973685821852684,
      "loss": 5.389,
      "mean_token_accuracy": 0.16972415894269943,
      "num_tokens": 12531783.0,
      "step": 7215
    },
    {
      "entropy": 5.604793548583984,
      "epoch": 0.5617584127601634,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973643541116528,
      "loss": 5.4463,
      "mean_token_accuracy": 0.16353272944688796,
      "num_tokens": 12541768.0,
      "step": 7220
    },
    {
      "entropy": 5.689733552932739,
      "epoch": 0.562147442131881,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497360122664012,
      "loss": 5.4659,
      "mean_token_accuracy": 0.16686318814754486,
      "num_tokens": 12550684.0,
      "step": 7225
    },
    {
      "entropy": 5.634374666213989,
      "epoch": 0.5625364715035985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973558878424104,
      "loss": 5.4107,
      "mean_token_accuracy": 0.16748277097940445,
      "num_tokens": 12559948.0,
      "step": 7230
    },
    {
      "entropy": 5.570274305343628,
      "epoch": 0.5629255008753161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004973516496469119,
      "loss": 5.3538,
      "mean_token_accuracy": 0.17910895943641664,
      "num_tokens": 12568681.0,
      "step": 7235
    },
    {
      "entropy": 5.688456439971924,
      "epoch": 0.5633145302470336,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973474080775811,
      "loss": 5.4826,
      "mean_token_accuracy": 0.1591241553425789,
      "num_tokens": 12577724.0,
      "step": 7240
    },
    {
      "entropy": 5.694258785247802,
      "epoch": 0.5637035596187512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973431631344824,
      "loss": 5.4048,
      "mean_token_accuracy": 0.16929993778467178,
      "num_tokens": 12585994.0,
      "step": 7245
    },
    {
      "entropy": 5.627398109436035,
      "epoch": 0.5640925889904688,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049733891481768,
      "loss": 5.4558,
      "mean_token_accuracy": 0.16533113718032838,
      "num_tokens": 12595525.0,
      "step": 7250
    },
    {
      "entropy": 5.62390513420105,
      "epoch": 0.5644816183621864,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004973346631272384,
      "loss": 5.2904,
      "mean_token_accuracy": 0.17755208760499955,
      "num_tokens": 12603645.0,
      "step": 7255
    },
    {
      "entropy": 5.567868185043335,
      "epoch": 0.5648706477339039,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973304080632222,
      "loss": 5.3675,
      "mean_token_accuracy": 0.1723654091358185,
      "num_tokens": 12612165.0,
      "step": 7260
    },
    {
      "entropy": 5.58600640296936,
      "epoch": 0.5652596771056215,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497326149625696,
      "loss": 5.2778,
      "mean_token_accuracy": 0.18184001594781876,
      "num_tokens": 12620678.0,
      "step": 7265
    },
    {
      "entropy": 5.679574966430664,
      "epoch": 0.565648706477339,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004973218878147244,
      "loss": 5.4693,
      "mean_token_accuracy": 0.173343163728714,
      "num_tokens": 12628745.0,
      "step": 7270
    },
    {
      "entropy": 5.6358977317810055,
      "epoch": 0.5660377358490566,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004973176226303719,
      "loss": 5.5254,
      "mean_token_accuracy": 0.16010626703500747,
      "num_tokens": 12637448.0,
      "step": 7275
    },
    {
      "entropy": 5.626923656463623,
      "epoch": 0.5664267652207742,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004973133540727033,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1741466835141182,
      "num_tokens": 12645804.0,
      "step": 7280
    },
    {
      "entropy": 5.602758455276489,
      "epoch": 0.5668157945924918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004973090821417835,
      "loss": 5.3544,
      "mean_token_accuracy": 0.173034206032753,
      "num_tokens": 12654156.0,
      "step": 7285
    },
    {
      "entropy": 5.572364473342896,
      "epoch": 0.5672048239642093,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973048068376772,
      "loss": 5.3077,
      "mean_token_accuracy": 0.17897556275129317,
      "num_tokens": 12662271.0,
      "step": 7290
    },
    {
      "entropy": 5.708192586898804,
      "epoch": 0.5675938533359268,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004973005281604492,
      "loss": 5.4874,
      "mean_token_accuracy": 0.16635391861200333,
      "num_tokens": 12671016.0,
      "step": 7295
    },
    {
      "entropy": 5.639117336273193,
      "epoch": 0.5679828827076444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972962461101646,
      "loss": 5.4305,
      "mean_token_accuracy": 0.1690814435482025,
      "num_tokens": 12679593.0,
      "step": 7300
    },
    {
      "entropy": 5.6017145156860355,
      "epoch": 0.568371912079362,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972919606868883,
      "loss": 5.3721,
      "mean_token_accuracy": 0.1766174241900444,
      "num_tokens": 12688715.0,
      "step": 7305
    },
    {
      "entropy": 5.602231359481811,
      "epoch": 0.5687609414510796,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972876718906852,
      "loss": 5.3866,
      "mean_token_accuracy": 0.1696537181735039,
      "num_tokens": 12698078.0,
      "step": 7310
    },
    {
      "entropy": 5.563684892654419,
      "epoch": 0.5691499708227972,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972833797216206,
      "loss": 5.3618,
      "mean_token_accuracy": 0.1765790119767189,
      "num_tokens": 12706375.0,
      "step": 7315
    },
    {
      "entropy": 5.715937089920044,
      "epoch": 0.5695390001945146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972790841797595,
      "loss": 5.5603,
      "mean_token_accuracy": 0.16448500901460647,
      "num_tokens": 12715727.0,
      "step": 7320
    },
    {
      "entropy": 5.664407587051391,
      "epoch": 0.5699280295662322,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497274785265167,
      "loss": 5.4516,
      "mean_token_accuracy": 0.1611158773303032,
      "num_tokens": 12725565.0,
      "step": 7325
    },
    {
      "entropy": 5.683152484893799,
      "epoch": 0.5703170589379498,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972704829779086,
      "loss": 5.4712,
      "mean_token_accuracy": 0.1761600434780121,
      "num_tokens": 12733498.0,
      "step": 7330
    },
    {
      "entropy": 5.556344842910766,
      "epoch": 0.5707060883096674,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972661773180492,
      "loss": 5.3119,
      "mean_token_accuracy": 0.1757061630487442,
      "num_tokens": 12741635.0,
      "step": 7335
    },
    {
      "entropy": 5.554010105133057,
      "epoch": 0.571095117681385,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972618682856545,
      "loss": 5.3269,
      "mean_token_accuracy": 0.1744554191827774,
      "num_tokens": 12750973.0,
      "step": 7340
    },
    {
      "entropy": 5.582947015762329,
      "epoch": 0.5714841470531025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004972575558807896,
      "loss": 5.3894,
      "mean_token_accuracy": 0.17799339741468428,
      "num_tokens": 12759075.0,
      "step": 7345
    },
    {
      "entropy": 5.646500492095948,
      "epoch": 0.57187317642482,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004972532401035201,
      "loss": 5.3664,
      "mean_token_accuracy": 0.17144054025411606,
      "num_tokens": 12767485.0,
      "step": 7350
    },
    {
      "entropy": 5.687675714492798,
      "epoch": 0.5722622057965376,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972489209539114,
      "loss": 5.3616,
      "mean_token_accuracy": 0.17879957854747772,
      "num_tokens": 12775128.0,
      "step": 7355
    },
    {
      "entropy": 5.556820583343506,
      "epoch": 0.5726512351682552,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972445984320292,
      "loss": 5.2769,
      "mean_token_accuracy": 0.17778964787721635,
      "num_tokens": 12783243.0,
      "step": 7360
    },
    {
      "entropy": 5.63396053314209,
      "epoch": 0.5730402645399728,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972402725379388,
      "loss": 5.4811,
      "mean_token_accuracy": 0.17174470275640488,
      "num_tokens": 12791613.0,
      "step": 7365
    },
    {
      "entropy": 5.671663236618042,
      "epoch": 0.5734292939116903,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497235943271706,
      "loss": 5.4198,
      "mean_token_accuracy": 0.1699995830655098,
      "num_tokens": 12799817.0,
      "step": 7370
    },
    {
      "entropy": 5.642856693267822,
      "epoch": 0.5738183232834079,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972316106333966,
      "loss": 5.4209,
      "mean_token_accuracy": 0.16815965324640275,
      "num_tokens": 12808994.0,
      "step": 7375
    },
    {
      "entropy": 5.642169857025147,
      "epoch": 0.5742073526551255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004972272746230761,
      "loss": 5.3904,
      "mean_token_accuracy": 0.16785614788532258,
      "num_tokens": 12817191.0,
      "step": 7380
    },
    {
      "entropy": 5.608738899230957,
      "epoch": 0.574596382026843,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972229352408104,
      "loss": 5.35,
      "mean_token_accuracy": 0.1734204888343811,
      "num_tokens": 12825773.0,
      "step": 7385
    },
    {
      "entropy": 5.678319597244263,
      "epoch": 0.5749854113985606,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972185924866655,
      "loss": 5.4246,
      "mean_token_accuracy": 0.1701473280787468,
      "num_tokens": 12834489.0,
      "step": 7390
    },
    {
      "entropy": 5.6620612144470215,
      "epoch": 0.5753744407702781,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972142463607071,
      "loss": 5.4356,
      "mean_token_accuracy": 0.17158037573099136,
      "num_tokens": 12843179.0,
      "step": 7395
    },
    {
      "entropy": 5.712574434280396,
      "epoch": 0.5757634701419957,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972098968630012,
      "loss": 5.4387,
      "mean_token_accuracy": 0.1631328746676445,
      "num_tokens": 12851243.0,
      "step": 7400
    },
    {
      "entropy": 5.585179996490479,
      "epoch": 0.5761524995137133,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972055439936137,
      "loss": 5.477,
      "mean_token_accuracy": 0.17327615469694138,
      "num_tokens": 12860267.0,
      "step": 7405
    },
    {
      "entropy": 5.654938220977783,
      "epoch": 0.5765415288854309,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972011877526109,
      "loss": 5.4859,
      "mean_token_accuracy": 0.1667654201388359,
      "num_tokens": 12870022.0,
      "step": 7410
    },
    {
      "entropy": 5.646070671081543,
      "epoch": 0.5769305582571485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971968281400587,
      "loss": 5.4501,
      "mean_token_accuracy": 0.1635710835456848,
      "num_tokens": 12879227.0,
      "step": 7415
    },
    {
      "entropy": 5.667190790176392,
      "epoch": 0.5773195876288659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971924651560233,
      "loss": 5.415,
      "mean_token_accuracy": 0.16541746705770494,
      "num_tokens": 12888180.0,
      "step": 7420
    },
    {
      "entropy": 5.6017412662506105,
      "epoch": 0.5777086170005835,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004971880988005709,
      "loss": 5.3977,
      "mean_token_accuracy": 0.17384001314640046,
      "num_tokens": 12896767.0,
      "step": 7425
    },
    {
      "entropy": 5.604390096664429,
      "epoch": 0.5780976463723011,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971837290737677,
      "loss": 5.2896,
      "mean_token_accuracy": 0.18285758793354034,
      "num_tokens": 12904766.0,
      "step": 7430
    },
    {
      "entropy": 5.549197387695313,
      "epoch": 0.5784866757440187,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004971793559756801,
      "loss": 5.371,
      "mean_token_accuracy": 0.16989658921957015,
      "num_tokens": 12913808.0,
      "step": 7435
    },
    {
      "entropy": 5.551879167556763,
      "epoch": 0.5788757051157363,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971749795063744,
      "loss": 5.3784,
      "mean_token_accuracy": 0.16979086995124817,
      "num_tokens": 12923103.0,
      "step": 7440
    },
    {
      "entropy": 5.59029188156128,
      "epoch": 0.5792647344874537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497170599665917,
      "loss": 5.3343,
      "mean_token_accuracy": 0.17453575134277344,
      "num_tokens": 12931168.0,
      "step": 7445
    },
    {
      "entropy": 5.5925538539886475,
      "epoch": 0.5796537638591713,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971662164543743,
      "loss": 5.4042,
      "mean_token_accuracy": 0.17637941390275955,
      "num_tokens": 12939875.0,
      "step": 7450
    },
    {
      "entropy": 5.599119186401367,
      "epoch": 0.5800427932308889,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004971618298718131,
      "loss": 5.325,
      "mean_token_accuracy": 0.17663128972053527,
      "num_tokens": 12947883.0,
      "step": 7455
    },
    {
      "entropy": 5.71734848022461,
      "epoch": 0.5804318226026065,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971574399182995,
      "loss": 5.5659,
      "mean_token_accuracy": 0.16337292790412902,
      "num_tokens": 12956940.0,
      "step": 7460
    },
    {
      "entropy": 5.600492000579834,
      "epoch": 0.5808208519743241,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004971530465939005,
      "loss": 5.3008,
      "mean_token_accuracy": 0.1726179912686348,
      "num_tokens": 12965003.0,
      "step": 7465
    },
    {
      "entropy": 5.57530517578125,
      "epoch": 0.5812098813460417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004971486498986825,
      "loss": 5.3198,
      "mean_token_accuracy": 0.1755088374018669,
      "num_tokens": 12973323.0,
      "step": 7470
    },
    {
      "entropy": 5.580819034576416,
      "epoch": 0.5815989107177592,
      "grad_norm": 1.0,
      "learning_rate": 0.0004971442498327124,
      "loss": 5.374,
      "mean_token_accuracy": 0.17628689855337143,
      "num_tokens": 12982654.0,
      "step": 7475
    },
    {
      "entropy": 5.610224342346191,
      "epoch": 0.5819879400894767,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004971398463960568,
      "loss": 5.3965,
      "mean_token_accuracy": 0.16887464374303818,
      "num_tokens": 12991781.0,
      "step": 7480
    },
    {
      "entropy": 5.694233083724976,
      "epoch": 0.5823769694611943,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971354395887827,
      "loss": 5.4907,
      "mean_token_accuracy": 0.16564130187034606,
      "num_tokens": 13000539.0,
      "step": 7485
    },
    {
      "entropy": 5.639818954467773,
      "epoch": 0.5827659988329119,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004971310294109569,
      "loss": 5.3575,
      "mean_token_accuracy": 0.17220927327871322,
      "num_tokens": 13008638.0,
      "step": 7490
    },
    {
      "entropy": 5.582267093658447,
      "epoch": 0.5831550282046295,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971266158626461,
      "loss": 5.3337,
      "mean_token_accuracy": 0.17774504721164702,
      "num_tokens": 13017217.0,
      "step": 7495
    },
    {
      "entropy": 5.579143905639649,
      "epoch": 0.583544057576347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971221989439177,
      "loss": 5.3991,
      "mean_token_accuracy": 0.1740799993276596,
      "num_tokens": 13025278.0,
      "step": 7500
    },
    {
      "entropy": 5.602803754806518,
      "epoch": 0.5839330869480646,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971177786548384,
      "loss": 5.4101,
      "mean_token_accuracy": 0.17012534886598588,
      "num_tokens": 13034484.0,
      "step": 7505
    },
    {
      "entropy": 5.68079833984375,
      "epoch": 0.5843221163197821,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971133549954753,
      "loss": 5.4796,
      "mean_token_accuracy": 0.16569795161485673,
      "num_tokens": 13042831.0,
      "step": 7510
    },
    {
      "entropy": 5.650001335144043,
      "epoch": 0.5847111456914997,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004971089279658954,
      "loss": 5.4474,
      "mean_token_accuracy": 0.16818668842315673,
      "num_tokens": 13051614.0,
      "step": 7515
    },
    {
      "entropy": 5.640414476394653,
      "epoch": 0.5851001750632173,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004971044975661663,
      "loss": 5.4016,
      "mean_token_accuracy": 0.17426666915416716,
      "num_tokens": 13060179.0,
      "step": 7520
    },
    {
      "entropy": 5.551918649673462,
      "epoch": 0.5854892044349348,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971000637963549,
      "loss": 5.329,
      "mean_token_accuracy": 0.18348958790302278,
      "num_tokens": 13068328.0,
      "step": 7525
    },
    {
      "entropy": 5.5161439895629885,
      "epoch": 0.5858782338066524,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970956266565285,
      "loss": 5.3816,
      "mean_token_accuracy": 0.17077172845602034,
      "num_tokens": 13076879.0,
      "step": 7530
    },
    {
      "entropy": 5.701922178268433,
      "epoch": 0.58626726317837,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970911861467544,
      "loss": 5.402,
      "mean_token_accuracy": 0.1686718076467514,
      "num_tokens": 13085123.0,
      "step": 7535
    },
    {
      "entropy": 5.559877395629883,
      "epoch": 0.5866562925500876,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970867422671002,
      "loss": 5.3854,
      "mean_token_accuracy": 0.17493418753147125,
      "num_tokens": 13093884.0,
      "step": 7540
    },
    {
      "entropy": 5.691018915176391,
      "epoch": 0.5870453219218051,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497082295017633,
      "loss": 5.4406,
      "mean_token_accuracy": 0.1709304377436638,
      "num_tokens": 13102581.0,
      "step": 7545
    },
    {
      "entropy": 5.638440179824829,
      "epoch": 0.5874343512935226,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970778443984206,
      "loss": 5.4096,
      "mean_token_accuracy": 0.17096369117498397,
      "num_tokens": 13111307.0,
      "step": 7550
    },
    {
      "entropy": 5.627034854888916,
      "epoch": 0.5878233806652402,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970733904095303,
      "loss": 5.4292,
      "mean_token_accuracy": 0.1729509636759758,
      "num_tokens": 13119279.0,
      "step": 7555
    },
    {
      "entropy": 5.680221939086914,
      "epoch": 0.5882124100369578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970689330510298,
      "loss": 5.4427,
      "mean_token_accuracy": 0.17425399124622346,
      "num_tokens": 13127753.0,
      "step": 7560
    },
    {
      "entropy": 5.631756019592285,
      "epoch": 0.5886014394086754,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970644723229868,
      "loss": 5.4736,
      "mean_token_accuracy": 0.16918028891086578,
      "num_tokens": 13136881.0,
      "step": 7565
    },
    {
      "entropy": 5.592969179153442,
      "epoch": 0.588990468780393,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970600082254687,
      "loss": 5.3553,
      "mean_token_accuracy": 0.1674942895770073,
      "num_tokens": 13145291.0,
      "step": 7570
    },
    {
      "entropy": 5.5917394161224365,
      "epoch": 0.5893794981521104,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970555407585436,
      "loss": 5.3184,
      "mean_token_accuracy": 0.16939782500267028,
      "num_tokens": 13153560.0,
      "step": 7575
    },
    {
      "entropy": 5.623266267776489,
      "epoch": 0.589768527523828,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497051069922279,
      "loss": 5.3725,
      "mean_token_accuracy": 0.1684000313282013,
      "num_tokens": 13162304.0,
      "step": 7580
    },
    {
      "entropy": 5.644482421875,
      "epoch": 0.5901575568955456,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970465957167429,
      "loss": 5.4001,
      "mean_token_accuracy": 0.1760484278202057,
      "num_tokens": 13171404.0,
      "step": 7585
    },
    {
      "entropy": 5.709658241271972,
      "epoch": 0.5905465862672632,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497042118142003,
      "loss": 5.4724,
      "mean_token_accuracy": 0.17499999850988388,
      "num_tokens": 13180001.0,
      "step": 7590
    },
    {
      "entropy": 5.655188035964966,
      "epoch": 0.5909356156389808,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970376371981275,
      "loss": 5.4073,
      "mean_token_accuracy": 0.17146350145339967,
      "num_tokens": 13189749.0,
      "step": 7595
    },
    {
      "entropy": 5.583410549163818,
      "epoch": 0.5913246450106983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970331528851842,
      "loss": 5.3954,
      "mean_token_accuracy": 0.17423295825719834,
      "num_tokens": 13199169.0,
      "step": 7600
    },
    {
      "entropy": 5.586973857879639,
      "epoch": 0.5917136743824158,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970286652032412,
      "loss": 5.3387,
      "mean_token_accuracy": 0.17489242255687715,
      "num_tokens": 13207014.0,
      "step": 7605
    },
    {
      "entropy": 5.605558156967163,
      "epoch": 0.5921027037541334,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970241741523667,
      "loss": 5.2923,
      "mean_token_accuracy": 0.179923477768898,
      "num_tokens": 13215333.0,
      "step": 7610
    },
    {
      "entropy": 5.582746839523315,
      "epoch": 0.592491733125851,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970196797326286,
      "loss": 5.4311,
      "mean_token_accuracy": 0.16824073791503907,
      "num_tokens": 13223740.0,
      "step": 7615
    },
    {
      "entropy": 5.560179328918457,
      "epoch": 0.5928807624975686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970151819440955,
      "loss": 5.3647,
      "mean_token_accuracy": 0.17508744299411774,
      "num_tokens": 13232065.0,
      "step": 7620
    },
    {
      "entropy": 5.6420036315917965,
      "epoch": 0.5932697918692861,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970106807868351,
      "loss": 5.35,
      "mean_token_accuracy": 0.17854661941528321,
      "num_tokens": 13241171.0,
      "step": 7625
    },
    {
      "entropy": 5.522149991989136,
      "epoch": 0.5936588212410037,
      "grad_norm": 1.0,
      "learning_rate": 0.000497006176260916,
      "loss": 5.2764,
      "mean_token_accuracy": 0.18158416748046874,
      "num_tokens": 13249553.0,
      "step": 7630
    },
    {
      "entropy": 5.586422872543335,
      "epoch": 0.5940478506127213,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970016683664068,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1729905426502228,
      "num_tokens": 13258585.0,
      "step": 7635
    },
    {
      "entropy": 5.614265632629395,
      "epoch": 0.5944368799844388,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969971571033754,
      "loss": 5.4276,
      "mean_token_accuracy": 0.17271430492401124,
      "num_tokens": 13267317.0,
      "step": 7640
    },
    {
      "entropy": 5.595805406570435,
      "epoch": 0.5948259093561564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969926424718906,
      "loss": 5.2315,
      "mean_token_accuracy": 0.17310887724161148,
      "num_tokens": 13275705.0,
      "step": 7645
    },
    {
      "entropy": 5.58509931564331,
      "epoch": 0.5952149387278739,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004969881244720207,
      "loss": 5.3738,
      "mean_token_accuracy": 0.17260442823171615,
      "num_tokens": 13284046.0,
      "step": 7650
    },
    {
      "entropy": 5.564942646026611,
      "epoch": 0.5956039680995915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004969836031038343,
      "loss": 5.3487,
      "mean_token_accuracy": 0.17941556870937347,
      "num_tokens": 13292451.0,
      "step": 7655
    },
    {
      "entropy": 5.695221996307373,
      "epoch": 0.5959929974713091,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004969790783674002,
      "loss": 5.4852,
      "mean_token_accuracy": 0.1700387567281723,
      "num_tokens": 13300788.0,
      "step": 7660
    },
    {
      "entropy": 5.654481363296509,
      "epoch": 0.5963820268430267,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969745502627868,
      "loss": 5.397,
      "mean_token_accuracy": 0.17259870916604997,
      "num_tokens": 13309235.0,
      "step": 7665
    },
    {
      "entropy": 5.555170011520386,
      "epoch": 0.5967710562147442,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969700187900629,
      "loss": 5.307,
      "mean_token_accuracy": 0.17510560154914856,
      "num_tokens": 13317331.0,
      "step": 7670
    },
    {
      "entropy": 5.562815141677857,
      "epoch": 0.5971600855864618,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969654839492973,
      "loss": 5.3472,
      "mean_token_accuracy": 0.16236402094364166,
      "num_tokens": 13327296.0,
      "step": 7675
    },
    {
      "entropy": 5.660052680969239,
      "epoch": 0.5975491149581793,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969609457405588,
      "loss": 5.4045,
      "mean_token_accuracy": 0.16961953341960906,
      "num_tokens": 13336650.0,
      "step": 7680
    },
    {
      "entropy": 5.620706462860108,
      "epoch": 0.5979381443298969,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969564041639163,
      "loss": 5.3521,
      "mean_token_accuracy": 0.17291512042284013,
      "num_tokens": 13345353.0,
      "step": 7685
    },
    {
      "entropy": 5.628172159194946,
      "epoch": 0.5983271737016145,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969518592194386,
      "loss": 5.3714,
      "mean_token_accuracy": 0.1745850458741188,
      "num_tokens": 13354302.0,
      "step": 7690
    },
    {
      "entropy": 5.6162646293640135,
      "epoch": 0.5987162030733321,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969473109071946,
      "loss": 5.314,
      "mean_token_accuracy": 0.17899066358804702,
      "num_tokens": 13362950.0,
      "step": 7695
    },
    {
      "entropy": 5.625281095504761,
      "epoch": 0.5991052324450497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004969427592272535,
      "loss": 5.4219,
      "mean_token_accuracy": 0.16785797625780105,
      "num_tokens": 13371383.0,
      "step": 7700
    },
    {
      "entropy": 5.638612365722656,
      "epoch": 0.5994942618167671,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969382041796843,
      "loss": 5.3742,
      "mean_token_accuracy": 0.18112112283706666,
      "num_tokens": 13379767.0,
      "step": 7705
    },
    {
      "entropy": 5.630096673965454,
      "epoch": 0.5998832911884847,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496933645764556,
      "loss": 5.3864,
      "mean_token_accuracy": 0.16946639716625214,
      "num_tokens": 13387970.0,
      "step": 7710
    },
    {
      "entropy": 5.543959760665894,
      "epoch": 0.6002723205602023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969290839819381,
      "loss": 5.3336,
      "mean_token_accuracy": 0.1790908545255661,
      "num_tokens": 13396605.0,
      "step": 7715
    },
    {
      "entropy": 5.550653791427612,
      "epoch": 0.6006613499319199,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969245188318994,
      "loss": 5.2794,
      "mean_token_accuracy": 0.18368381559848784,
      "num_tokens": 13406028.0,
      "step": 7720
    },
    {
      "entropy": 5.612054634094238,
      "epoch": 0.6010503793036375,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004969199503145093,
      "loss": 5.3223,
      "mean_token_accuracy": 0.181209759414196,
      "num_tokens": 13414927.0,
      "step": 7725
    },
    {
      "entropy": 5.531005764007569,
      "epoch": 0.601439408675355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969153784298374,
      "loss": 5.284,
      "mean_token_accuracy": 0.1768856644630432,
      "num_tokens": 13423020.0,
      "step": 7730
    },
    {
      "entropy": 5.509295701980591,
      "epoch": 0.6018284380470725,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969108031779527,
      "loss": 5.2792,
      "mean_token_accuracy": 0.17936475723981857,
      "num_tokens": 13431514.0,
      "step": 7735
    },
    {
      "entropy": 5.622158241271973,
      "epoch": 0.6022174674187901,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969062245589247,
      "loss": 5.451,
      "mean_token_accuracy": 0.1655518352985382,
      "num_tokens": 13440333.0,
      "step": 7740
    },
    {
      "entropy": 5.601479339599609,
      "epoch": 0.6026064967905077,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969016425728231,
      "loss": 5.3896,
      "mean_token_accuracy": 0.17400994896888733,
      "num_tokens": 13448841.0,
      "step": 7745
    },
    {
      "entropy": 5.620350360870361,
      "epoch": 0.6029955261622253,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004968970572197172,
      "loss": 5.4691,
      "mean_token_accuracy": 0.16713852137327195,
      "num_tokens": 13457846.0,
      "step": 7750
    },
    {
      "entropy": 5.697343492507935,
      "epoch": 0.6033845555339428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968924684996765,
      "loss": 5.4593,
      "mean_token_accuracy": 0.16241354048252105,
      "num_tokens": 13466242.0,
      "step": 7755
    },
    {
      "entropy": 5.621889352798462,
      "epoch": 0.6037735849056604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496887876412771,
      "loss": 5.4069,
      "mean_token_accuracy": 0.16790285110473632,
      "num_tokens": 13474639.0,
      "step": 7760
    },
    {
      "entropy": 5.529103708267212,
      "epoch": 0.6041626142773779,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968832809590701,
      "loss": 5.2877,
      "mean_token_accuracy": 0.1749562978744507,
      "num_tokens": 13483673.0,
      "step": 7765
    },
    {
      "entropy": 5.654218864440918,
      "epoch": 0.6045516436490955,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004968786821386435,
      "loss": 5.403,
      "mean_token_accuracy": 0.16918199509382248,
      "num_tokens": 13493001.0,
      "step": 7770
    },
    {
      "entropy": 5.737928676605224,
      "epoch": 0.6049406730208131,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004968740799515611,
      "loss": 5.4196,
      "mean_token_accuracy": 0.17329576909542083,
      "num_tokens": 13500878.0,
      "step": 7775
    },
    {
      "entropy": 5.644543170928955,
      "epoch": 0.6053297023925306,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968694743978926,
      "loss": 5.4742,
      "mean_token_accuracy": 0.16776228249073027,
      "num_tokens": 13509945.0,
      "step": 7780
    },
    {
      "entropy": 5.5481706142425535,
      "epoch": 0.6057187317642482,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968648654777081,
      "loss": 5.4034,
      "mean_token_accuracy": 0.16965004354715346,
      "num_tokens": 13520024.0,
      "step": 7785
    },
    {
      "entropy": 5.619679021835327,
      "epoch": 0.6061077611359658,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968602531910773,
      "loss": 5.4879,
      "mean_token_accuracy": 0.1689732030034065,
      "num_tokens": 13529136.0,
      "step": 7790
    },
    {
      "entropy": 5.604547309875488,
      "epoch": 0.6064967905076833,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968556375380704,
      "loss": 5.3067,
      "mean_token_accuracy": 0.17450194507837297,
      "num_tokens": 13537001.0,
      "step": 7795
    },
    {
      "entropy": 5.6026702404022215,
      "epoch": 0.6068858198794009,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968510185187571,
      "loss": 5.2886,
      "mean_token_accuracy": 0.17515256106853486,
      "num_tokens": 13545520.0,
      "step": 7800
    },
    {
      "entropy": 5.569001913070679,
      "epoch": 0.6072748492511184,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496846396133208,
      "loss": 5.3665,
      "mean_token_accuracy": 0.1716013804078102,
      "num_tokens": 13554770.0,
      "step": 7805
    },
    {
      "entropy": 5.668507719039917,
      "epoch": 0.607663878622836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968417703814928,
      "loss": 5.4092,
      "mean_token_accuracy": 0.1714569166302681,
      "num_tokens": 13562797.0,
      "step": 7810
    },
    {
      "entropy": 5.597845935821534,
      "epoch": 0.6080529079945536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968371412636818,
      "loss": 5.3629,
      "mean_token_accuracy": 0.17308429032564163,
      "num_tokens": 13572017.0,
      "step": 7815
    },
    {
      "entropy": 5.6359875202178955,
      "epoch": 0.6084419373662712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968325087798453,
      "loss": 5.3776,
      "mean_token_accuracy": 0.1751322031021118,
      "num_tokens": 13580245.0,
      "step": 7820
    },
    {
      "entropy": 5.610501098632812,
      "epoch": 0.6088309667379888,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968278729300536,
      "loss": 5.3483,
      "mean_token_accuracy": 0.1653791770339012,
      "num_tokens": 13588719.0,
      "step": 7825
    },
    {
      "entropy": 5.660403871536255,
      "epoch": 0.6092199961097062,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968232337143769,
      "loss": 5.3892,
      "mean_token_accuracy": 0.17397119104862213,
      "num_tokens": 13597345.0,
      "step": 7830
    },
    {
      "entropy": 5.665544080734253,
      "epoch": 0.6096090254814238,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968185911328858,
      "loss": 5.4156,
      "mean_token_accuracy": 0.17118645012378692,
      "num_tokens": 13605468.0,
      "step": 7835
    },
    {
      "entropy": 5.653286933898926,
      "epoch": 0.6099980548531414,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004968139451856506,
      "loss": 5.4297,
      "mean_token_accuracy": 0.16750675886869432,
      "num_tokens": 13614085.0,
      "step": 7840
    },
    {
      "entropy": 5.5557215213775635,
      "epoch": 0.610387084224859,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968092958727419,
      "loss": 5.2759,
      "mean_token_accuracy": 0.17956970781087875,
      "num_tokens": 13622370.0,
      "step": 7845
    },
    {
      "entropy": 5.6002819538116455,
      "epoch": 0.6107761135965766,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049680464319423,
      "loss": 5.3743,
      "mean_token_accuracy": 0.18000267893075944,
      "num_tokens": 13630615.0,
      "step": 7850
    },
    {
      "entropy": 5.631160593032837,
      "epoch": 0.6111651429682942,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004967999871501858,
      "loss": 5.2997,
      "mean_token_accuracy": 0.18156272917985916,
      "num_tokens": 13638461.0,
      "step": 7855
    },
    {
      "entropy": 5.560482788085937,
      "epoch": 0.6115541723400116,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967953277406798,
      "loss": 5.4074,
      "mean_token_accuracy": 0.1724785551428795,
      "num_tokens": 13647266.0,
      "step": 7860
    },
    {
      "entropy": 5.554763126373291,
      "epoch": 0.6119432017117292,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004967906649657828,
      "loss": 5.3939,
      "mean_token_accuracy": 0.17485640347003936,
      "num_tokens": 13655930.0,
      "step": 7865
    },
    {
      "entropy": 5.644390821456909,
      "epoch": 0.6123322310834468,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967859988255655,
      "loss": 5.3716,
      "mean_token_accuracy": 0.17199483066797255,
      "num_tokens": 13664427.0,
      "step": 7870
    },
    {
      "entropy": 5.62963171005249,
      "epoch": 0.6127212604551644,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004967813293200985,
      "loss": 5.4058,
      "mean_token_accuracy": 0.16890444457530976,
      "num_tokens": 13673120.0,
      "step": 7875
    },
    {
      "entropy": 5.534495687484741,
      "epoch": 0.613110289826882,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967766564494529,
      "loss": 5.2386,
      "mean_token_accuracy": 0.18175663352012633,
      "num_tokens": 13681918.0,
      "step": 7880
    },
    {
      "entropy": 5.55442247390747,
      "epoch": 0.6134993191985995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967719802136996,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1815569058060646,
      "num_tokens": 13689904.0,
      "step": 7885
    },
    {
      "entropy": 5.593478679656982,
      "epoch": 0.613888348570317,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004967673006129094,
      "loss": 5.3508,
      "mean_token_accuracy": 0.17641768157482146,
      "num_tokens": 13697861.0,
      "step": 7890
    },
    {
      "entropy": 5.706309461593628,
      "epoch": 0.6142773779420346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004967626176471536,
      "loss": 5.507,
      "mean_token_accuracy": 0.17445981055498122,
      "num_tokens": 13706058.0,
      "step": 7895
    },
    {
      "entropy": 5.639518213272095,
      "epoch": 0.6146664073137522,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967579313165028,
      "loss": 5.4266,
      "mean_token_accuracy": 0.17015932351350785,
      "num_tokens": 13715081.0,
      "step": 7900
    },
    {
      "entropy": 5.591048812866211,
      "epoch": 0.6150554366854698,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004967532416210284,
      "loss": 5.3097,
      "mean_token_accuracy": 0.17756472676992416,
      "num_tokens": 13723311.0,
      "step": 7905
    },
    {
      "entropy": 5.636404180526734,
      "epoch": 0.6154444660571873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967485485608016,
      "loss": 5.3424,
      "mean_token_accuracy": 0.1731407791376114,
      "num_tokens": 13731737.0,
      "step": 7910
    },
    {
      "entropy": 5.581039142608643,
      "epoch": 0.6158334954289049,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967438521358934,
      "loss": 5.3939,
      "mean_token_accuracy": 0.18245703727006912,
      "num_tokens": 13740255.0,
      "step": 7915
    },
    {
      "entropy": 5.601918697357178,
      "epoch": 0.6162225248006225,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004967391523463754,
      "loss": 5.3485,
      "mean_token_accuracy": 0.17710966467857361,
      "num_tokens": 13748656.0,
      "step": 7920
    },
    {
      "entropy": 5.601391744613648,
      "epoch": 0.61661155417234,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967344491923185,
      "loss": 5.3512,
      "mean_token_accuracy": 0.17627048045396804,
      "num_tokens": 13757539.0,
      "step": 7925
    },
    {
      "entropy": 5.63732876777649,
      "epoch": 0.6170005835440576,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967297426737943,
      "loss": 5.3672,
      "mean_token_accuracy": 0.17479392290115356,
      "num_tokens": 13766569.0,
      "step": 7930
    },
    {
      "entropy": 5.64272632598877,
      "epoch": 0.6173896129157751,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004967250327908742,
      "loss": 5.3842,
      "mean_token_accuracy": 0.17406178414821624,
      "num_tokens": 13775071.0,
      "step": 7935
    },
    {
      "entropy": 5.59261908531189,
      "epoch": 0.6177786422874927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004967203195436297,
      "loss": 5.3883,
      "mean_token_accuracy": 0.17029981613159179,
      "num_tokens": 13783319.0,
      "step": 7940
    },
    {
      "entropy": 5.504475259780884,
      "epoch": 0.6181676716592103,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967156029321322,
      "loss": 5.3205,
      "mean_token_accuracy": 0.17850407361984252,
      "num_tokens": 13792746.0,
      "step": 7945
    },
    {
      "entropy": 5.615712118148804,
      "epoch": 0.6185567010309279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004967108829564532,
      "loss": 5.2708,
      "mean_token_accuracy": 0.1812506452202797,
      "num_tokens": 13801632.0,
      "step": 7950
    },
    {
      "entropy": 5.6048766613006595,
      "epoch": 0.6189457304026454,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967061596166646,
      "loss": 5.4089,
      "mean_token_accuracy": 0.17564210891723633,
      "num_tokens": 13810527.0,
      "step": 7955
    },
    {
      "entropy": 5.635960817337036,
      "epoch": 0.6193347597743629,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967014329128378,
      "loss": 5.3415,
      "mean_token_accuracy": 0.17582118660211563,
      "num_tokens": 13819885.0,
      "step": 7960
    },
    {
      "entropy": 5.5821006298065186,
      "epoch": 0.6197237891460805,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966967028450448,
      "loss": 5.3487,
      "mean_token_accuracy": 0.17456280440092087,
      "num_tokens": 13828509.0,
      "step": 7965
    },
    {
      "entropy": 5.618126344680786,
      "epoch": 0.6201128185177981,
      "grad_norm": 1.0,
      "learning_rate": 0.000496691969413357,
      "loss": 5.3499,
      "mean_token_accuracy": 0.1717231258749962,
      "num_tokens": 13837196.0,
      "step": 7970
    },
    {
      "entropy": 5.672095489501953,
      "epoch": 0.6205018478895157,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004966872326178466,
      "loss": 5.4049,
      "mean_token_accuracy": 0.1715243563055992,
      "num_tokens": 13845849.0,
      "step": 7975
    },
    {
      "entropy": 5.6125000476837155,
      "epoch": 0.6208908772612333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966824924585851,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18061205595731736,
      "num_tokens": 13853430.0,
      "step": 7980
    },
    {
      "entropy": 5.624846410751343,
      "epoch": 0.6212799066329507,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966777489356448,
      "loss": 5.501,
      "mean_token_accuracy": 0.16109914481639862,
      "num_tokens": 13862520.0,
      "step": 7985
    },
    {
      "entropy": 5.689003944396973,
      "epoch": 0.6216689360046683,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966730020490975,
      "loss": 5.4134,
      "mean_token_accuracy": 0.16940829604864122,
      "num_tokens": 13871064.0,
      "step": 7990
    },
    {
      "entropy": 5.604720258712769,
      "epoch": 0.6220579653763859,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496668251799015,
      "loss": 5.3279,
      "mean_token_accuracy": 0.16955282390117646,
      "num_tokens": 13879952.0,
      "step": 7995
    },
    {
      "entropy": 5.669378137588501,
      "epoch": 0.6224469947481035,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004966634981854699,
      "loss": 5.4718,
      "mean_token_accuracy": 0.16619216203689574,
      "num_tokens": 13888054.0,
      "step": 8000
    },
    {
      "entropy": 5.682370376586914,
      "epoch": 0.6228360241198211,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966587412085339,
      "loss": 5.4597,
      "mean_token_accuracy": 0.16759533882141114,
      "num_tokens": 13897051.0,
      "step": 8005
    },
    {
      "entropy": 5.615130519866943,
      "epoch": 0.6232250534915386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966539808682793,
      "loss": 5.4087,
      "mean_token_accuracy": 0.16823190152645112,
      "num_tokens": 13906175.0,
      "step": 8010
    },
    {
      "entropy": 5.588666582107544,
      "epoch": 0.6236140828632561,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966492171647783,
      "loss": 5.3684,
      "mean_token_accuracy": 0.17251305282115936,
      "num_tokens": 13915283.0,
      "step": 8015
    },
    {
      "entropy": 5.611092948913575,
      "epoch": 0.6240031122349737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004966444500981032,
      "loss": 5.4001,
      "mean_token_accuracy": 0.17030985951423644,
      "num_tokens": 13924007.0,
      "step": 8020
    },
    {
      "entropy": 5.601291561126709,
      "epoch": 0.6243921416066913,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966396796683265,
      "loss": 5.3199,
      "mean_token_accuracy": 0.17982642948627472,
      "num_tokens": 13932694.0,
      "step": 8025
    },
    {
      "entropy": 5.662326431274414,
      "epoch": 0.6247811709784089,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004966349058755204,
      "loss": 5.4443,
      "mean_token_accuracy": 0.16524723023176194,
      "num_tokens": 13941817.0,
      "step": 8030
    },
    {
      "entropy": 5.623727893829345,
      "epoch": 0.6251702003501264,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966301287197573,
      "loss": 5.267,
      "mean_token_accuracy": 0.17870961129665375,
      "num_tokens": 13950413.0,
      "step": 8035
    },
    {
      "entropy": 5.67491660118103,
      "epoch": 0.625559229721844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004966253482011098,
      "loss": 5.4348,
      "mean_token_accuracy": 0.16312237679958344,
      "num_tokens": 13959415.0,
      "step": 8040
    },
    {
      "entropy": 5.560367059707642,
      "epoch": 0.6259482590935616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004966205643196503,
      "loss": 5.2852,
      "mean_token_accuracy": 0.178645259141922,
      "num_tokens": 13967889.0,
      "step": 8045
    },
    {
      "entropy": 5.473169136047363,
      "epoch": 0.6263372884652791,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966157770754516,
      "loss": 5.2764,
      "mean_token_accuracy": 0.17681868076324464,
      "num_tokens": 13976190.0,
      "step": 8050
    },
    {
      "entropy": 5.687845706939697,
      "epoch": 0.6267263178369967,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496610986468586,
      "loss": 5.4326,
      "mean_token_accuracy": 0.1652798607945442,
      "num_tokens": 13984940.0,
      "step": 8055
    },
    {
      "entropy": 5.671137046813965,
      "epoch": 0.6271153472087143,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004966061924991267,
      "loss": 5.4124,
      "mean_token_accuracy": 0.16456559300422668,
      "num_tokens": 13993740.0,
      "step": 8060
    },
    {
      "entropy": 5.642599296569824,
      "epoch": 0.6275043765804318,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496601395167146,
      "loss": 5.4491,
      "mean_token_accuracy": 0.17298204749822615,
      "num_tokens": 14002401.0,
      "step": 8065
    },
    {
      "entropy": 5.631506156921387,
      "epoch": 0.6278934059521494,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965965944727168,
      "loss": 5.4333,
      "mean_token_accuracy": 0.17514036148786544,
      "num_tokens": 14011197.0,
      "step": 8070
    },
    {
      "entropy": 5.577688980102539,
      "epoch": 0.628282435323867,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965917904159121,
      "loss": 5.3065,
      "mean_token_accuracy": 0.17777174413204194,
      "num_tokens": 14020344.0,
      "step": 8075
    },
    {
      "entropy": 5.550454330444336,
      "epoch": 0.6286714646955845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965869829968046,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17626356929540635,
      "num_tokens": 14029117.0,
      "step": 8080
    },
    {
      "entropy": 5.547106027603149,
      "epoch": 0.6290604940673021,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965821722154674,
      "loss": 5.2215,
      "mean_token_accuracy": 0.18387194871902465,
      "num_tokens": 14038395.0,
      "step": 8085
    },
    {
      "entropy": 5.582145690917969,
      "epoch": 0.6294495234390196,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004965773580719734,
      "loss": 5.3272,
      "mean_token_accuracy": 0.17480739206075668,
      "num_tokens": 14046983.0,
      "step": 8090
    },
    {
      "entropy": 5.510600614547729,
      "epoch": 0.6298385528107372,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965725405663957,
      "loss": 5.2224,
      "mean_token_accuracy": 0.18485897183418273,
      "num_tokens": 14056261.0,
      "step": 8095
    },
    {
      "entropy": 5.55096583366394,
      "epoch": 0.6302275821824548,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965677196988073,
      "loss": 5.3915,
      "mean_token_accuracy": 0.1706537574529648,
      "num_tokens": 14065371.0,
      "step": 8100
    },
    {
      "entropy": 5.64781346321106,
      "epoch": 0.6306166115541724,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965628954692815,
      "loss": 5.4435,
      "mean_token_accuracy": 0.1682393491268158,
      "num_tokens": 14074007.0,
      "step": 8105
    },
    {
      "entropy": 5.582576656341553,
      "epoch": 0.63100564092589,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965580678778913,
      "loss": 5.3122,
      "mean_token_accuracy": 0.17798371911048888,
      "num_tokens": 14083259.0,
      "step": 8110
    },
    {
      "entropy": 5.530705070495605,
      "epoch": 0.6313946702976074,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004965532369247102,
      "loss": 5.2977,
      "mean_token_accuracy": 0.1768136814236641,
      "num_tokens": 14090993.0,
      "step": 8115
    },
    {
      "entropy": 5.672610759735107,
      "epoch": 0.631783699669325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004965484026098111,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17126849442720413,
      "num_tokens": 14099722.0,
      "step": 8120
    },
    {
      "entropy": 5.69265775680542,
      "epoch": 0.6321727290410426,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965435649332679,
      "loss": 5.4181,
      "mean_token_accuracy": 0.17119095772504805,
      "num_tokens": 14108483.0,
      "step": 8125
    },
    {
      "entropy": 5.676124429702758,
      "epoch": 0.6325617584127602,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965387238951536,
      "loss": 5.4265,
      "mean_token_accuracy": 0.17131333351135253,
      "num_tokens": 14117187.0,
      "step": 8130
    },
    {
      "entropy": 5.570986270904541,
      "epoch": 0.6329507877844778,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004965338794955418,
      "loss": 5.2384,
      "mean_token_accuracy": 0.18281993120908738,
      "num_tokens": 14125424.0,
      "step": 8135
    },
    {
      "entropy": 5.582561111450195,
      "epoch": 0.6333398171561953,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965290317345061,
      "loss": 5.3342,
      "mean_token_accuracy": 0.17578710317611695,
      "num_tokens": 14133038.0,
      "step": 8140
    },
    {
      "entropy": 5.64760570526123,
      "epoch": 0.6337288465279128,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004965241806121198,
      "loss": 5.4718,
      "mean_token_accuracy": 0.16664536893367768,
      "num_tokens": 14141114.0,
      "step": 8145
    },
    {
      "entropy": 5.626469612121582,
      "epoch": 0.6341178758996304,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965193261284567,
      "loss": 5.3448,
      "mean_token_accuracy": 0.17470071762800216,
      "num_tokens": 14149681.0,
      "step": 8150
    },
    {
      "entropy": 5.643252897262573,
      "epoch": 0.634506905271348,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965144682835904,
      "loss": 5.3751,
      "mean_token_accuracy": 0.17150839567184448,
      "num_tokens": 14158510.0,
      "step": 8155
    },
    {
      "entropy": 5.612422370910645,
      "epoch": 0.6348959346430656,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965096070775946,
      "loss": 5.4212,
      "mean_token_accuracy": 0.1791756570339203,
      "num_tokens": 14166897.0,
      "step": 8160
    },
    {
      "entropy": 5.6259208679199215,
      "epoch": 0.6352849640147831,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965047425105431,
      "loss": 5.3637,
      "mean_token_accuracy": 0.17512688934803008,
      "num_tokens": 14174692.0,
      "step": 8165
    },
    {
      "entropy": 5.635723733901978,
      "epoch": 0.6356739933865007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004964998745825097,
      "loss": 5.3584,
      "mean_token_accuracy": 0.17260195165872574,
      "num_tokens": 14183779.0,
      "step": 8170
    },
    {
      "entropy": 5.611942434310913,
      "epoch": 0.6360630227582182,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964950032935682,
      "loss": 5.4123,
      "mean_token_accuracy": 0.17239698469638826,
      "num_tokens": 14191809.0,
      "step": 8175
    },
    {
      "entropy": 5.583318996429443,
      "epoch": 0.6364520521299358,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964901286437927,
      "loss": 5.3605,
      "mean_token_accuracy": 0.16911054253578187,
      "num_tokens": 14200652.0,
      "step": 8180
    },
    {
      "entropy": 5.60642819404602,
      "epoch": 0.6368410815016534,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004964852506332568,
      "loss": 5.3574,
      "mean_token_accuracy": 0.17232050895690917,
      "num_tokens": 14209694.0,
      "step": 8185
    },
    {
      "entropy": 5.61531982421875,
      "epoch": 0.6372301108733709,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496480369262035,
      "loss": 5.3981,
      "mean_token_accuracy": 0.17275434732437134,
      "num_tokens": 14218040.0,
      "step": 8190
    },
    {
      "entropy": 5.556940221786499,
      "epoch": 0.6376191402450885,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496475484530201,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1787574052810669,
      "num_tokens": 14225908.0,
      "step": 8195
    },
    {
      "entropy": 5.5955792427062985,
      "epoch": 0.6380081696168061,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964705964378292,
      "loss": 5.3469,
      "mean_token_accuracy": 0.16975195556879044,
      "num_tokens": 14234140.0,
      "step": 8200
    },
    {
      "entropy": 5.660179615020752,
      "epoch": 0.6383971989885237,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964657049849934,
      "loss": 5.4016,
      "mean_token_accuracy": 0.17422222793102266,
      "num_tokens": 14243590.0,
      "step": 8205
    },
    {
      "entropy": 5.640476036071777,
      "epoch": 0.6387862283602412,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964608101717681,
      "loss": 5.3899,
      "mean_token_accuracy": 0.17153700143098832,
      "num_tokens": 14252249.0,
      "step": 8210
    },
    {
      "entropy": 5.5549538135528564,
      "epoch": 0.6391752577319587,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004964559119982275,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1751054957509041,
      "num_tokens": 14261147.0,
      "step": 8215
    },
    {
      "entropy": 5.60577163696289,
      "epoch": 0.6395642871036763,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496451010464446,
      "loss": 5.381,
      "mean_token_accuracy": 0.17149227261543273,
      "num_tokens": 14270309.0,
      "step": 8220
    },
    {
      "entropy": 5.580153703689575,
      "epoch": 0.6399533164753939,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004964461055704978,
      "loss": 5.3534,
      "mean_token_accuracy": 0.17434841990470887,
      "num_tokens": 14279379.0,
      "step": 8225
    },
    {
      "entropy": 5.578339242935181,
      "epoch": 0.6403423458471115,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964411973164574,
      "loss": 5.3022,
      "mean_token_accuracy": 0.17278128117322922,
      "num_tokens": 14288136.0,
      "step": 8230
    },
    {
      "entropy": 5.548076820373535,
      "epoch": 0.6407313752188291,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964362857023993,
      "loss": 5.2367,
      "mean_token_accuracy": 0.1741736575961113,
      "num_tokens": 14296437.0,
      "step": 8235
    },
    {
      "entropy": 5.511275911331177,
      "epoch": 0.6411204045905465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000496431370728398,
      "loss": 5.2832,
      "mean_token_accuracy": 0.18518003970384597,
      "num_tokens": 14304861.0,
      "step": 8240
    },
    {
      "entropy": 5.499144077301025,
      "epoch": 0.6415094339622641,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004964264523945281,
      "loss": 5.3357,
      "mean_token_accuracy": 0.17871750742197037,
      "num_tokens": 14312861.0,
      "step": 8245
    },
    {
      "entropy": 5.495590162277222,
      "epoch": 0.6418984633339817,
      "grad_norm": 1.0,
      "learning_rate": 0.0004964215307008643,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1813568279147148,
      "num_tokens": 14321651.0,
      "step": 8250
    },
    {
      "entropy": 5.622010946273804,
      "epoch": 0.6422874927056993,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964166056474811,
      "loss": 5.3699,
      "mean_token_accuracy": 0.18560284674167632,
      "num_tokens": 14329234.0,
      "step": 8255
    },
    {
      "entropy": 5.614206838607788,
      "epoch": 0.6426765220774169,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964116772344534,
      "loss": 5.4891,
      "mean_token_accuracy": 0.17152170538902284,
      "num_tokens": 14337603.0,
      "step": 8260
    },
    {
      "entropy": 5.587452936172485,
      "epoch": 0.6430655514491345,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964067454618559,
      "loss": 5.3606,
      "mean_token_accuracy": 0.17309214770793915,
      "num_tokens": 14346634.0,
      "step": 8265
    },
    {
      "entropy": 5.642218828201294,
      "epoch": 0.6434545808208519,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964018103297634,
      "loss": 5.3903,
      "mean_token_accuracy": 0.1727163702249527,
      "num_tokens": 14355336.0,
      "step": 8270
    },
    {
      "entropy": 5.588451719284057,
      "epoch": 0.6438436101925695,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963968718382509,
      "loss": 5.3387,
      "mean_token_accuracy": 0.17309121489524842,
      "num_tokens": 14364434.0,
      "step": 8275
    },
    {
      "entropy": 5.589485788345337,
      "epoch": 0.6442326395642871,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963919299873931,
      "loss": 5.2902,
      "mean_token_accuracy": 0.1865809217095375,
      "num_tokens": 14372454.0,
      "step": 8280
    },
    {
      "entropy": 5.588171482086182,
      "epoch": 0.6446216689360047,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004963869847772653,
      "loss": 5.3273,
      "mean_token_accuracy": 0.17058257311582564,
      "num_tokens": 14381381.0,
      "step": 8285
    },
    {
      "entropy": 5.633420133590699,
      "epoch": 0.6450106983077223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004963820362079423,
      "loss": 5.4332,
      "mean_token_accuracy": 0.17251765131950378,
      "num_tokens": 14390853.0,
      "step": 8290
    },
    {
      "entropy": 5.5635377883911135,
      "epoch": 0.6453997276794398,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963770842794994,
      "loss": 5.2868,
      "mean_token_accuracy": 0.17360883802175522,
      "num_tokens": 14399897.0,
      "step": 8295
    },
    {
      "entropy": 5.515204811096192,
      "epoch": 0.6457887570511573,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004963721289920115,
      "loss": 5.3473,
      "mean_token_accuracy": 0.1733364999294281,
      "num_tokens": 14408192.0,
      "step": 8300
    },
    {
      "entropy": 5.562912654876709,
      "epoch": 0.6461777864228749,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963671703455538,
      "loss": 5.235,
      "mean_token_accuracy": 0.18256812691688537,
      "num_tokens": 14416004.0,
      "step": 8305
    },
    {
      "entropy": 5.563345527648925,
      "epoch": 0.6465668157945925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963622083402018,
      "loss": 5.2775,
      "mean_token_accuracy": 0.1796865537762642,
      "num_tokens": 14424573.0,
      "step": 8310
    },
    {
      "entropy": 5.556096458435059,
      "epoch": 0.6469558451663101,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963572429760305,
      "loss": 5.2634,
      "mean_token_accuracy": 0.18015891015529634,
      "num_tokens": 14432756.0,
      "step": 8315
    },
    {
      "entropy": 5.625596141815185,
      "epoch": 0.6473448745380276,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963522742531155,
      "loss": 5.3662,
      "mean_token_accuracy": 0.17505901157855988,
      "num_tokens": 14441695.0,
      "step": 8320
    },
    {
      "entropy": 5.582406759262085,
      "epoch": 0.6477339039097452,
      "grad_norm": 1.0,
      "learning_rate": 0.000496347302171532,
      "loss": 5.4448,
      "mean_token_accuracy": 0.1656550168991089,
      "num_tokens": 14451233.0,
      "step": 8325
    },
    {
      "entropy": 5.6259113311767575,
      "epoch": 0.6481229332814628,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963423267313554,
      "loss": 5.3415,
      "mean_token_accuracy": 0.17475675642490388,
      "num_tokens": 14459757.0,
      "step": 8330
    },
    {
      "entropy": 5.614810228347778,
      "epoch": 0.6485119626531803,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963373479326614,
      "loss": 5.4108,
      "mean_token_accuracy": 0.16576436012983323,
      "num_tokens": 14468308.0,
      "step": 8335
    },
    {
      "entropy": 5.608193254470825,
      "epoch": 0.6489009920248979,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963323657755255,
      "loss": 5.3428,
      "mean_token_accuracy": 0.18068965524435043,
      "num_tokens": 14476654.0,
      "step": 8340
    },
    {
      "entropy": 5.615989542007446,
      "epoch": 0.6492900213966154,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963273802600231,
      "loss": 5.4243,
      "mean_token_accuracy": 0.17392317503690718,
      "num_tokens": 14484960.0,
      "step": 8345
    },
    {
      "entropy": 5.67059440612793,
      "epoch": 0.649679050768333,
      "grad_norm": 1.0,
      "learning_rate": 0.00049632239138623,
      "loss": 5.4473,
      "mean_token_accuracy": 0.16484784781932832,
      "num_tokens": 14493708.0,
      "step": 8350
    },
    {
      "entropy": 5.6571794033050535,
      "epoch": 0.6500680801400506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963173991542219,
      "loss": 5.3834,
      "mean_token_accuracy": 0.1717427045106888,
      "num_tokens": 14502123.0,
      "step": 8355
    },
    {
      "entropy": 5.606250286102295,
      "epoch": 0.6504571095117682,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004963124035640746,
      "loss": 5.3403,
      "mean_token_accuracy": 0.17879369109869003,
      "num_tokens": 14510782.0,
      "step": 8360
    },
    {
      "entropy": 5.551813840866089,
      "epoch": 0.6508461388834857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963074046158637,
      "loss": 5.305,
      "mean_token_accuracy": 0.17476321160793304,
      "num_tokens": 14518560.0,
      "step": 8365
    },
    {
      "entropy": 5.608090591430664,
      "epoch": 0.6512351682552032,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963024023096652,
      "loss": 5.3185,
      "mean_token_accuracy": 0.17200247645378114,
      "num_tokens": 14527541.0,
      "step": 8370
    },
    {
      "entropy": 5.662448120117188,
      "epoch": 0.6516241976269208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004962973966455551,
      "loss": 5.3703,
      "mean_token_accuracy": 0.17151203900575637,
      "num_tokens": 14535590.0,
      "step": 8375
    },
    {
      "entropy": 5.5442280769348145,
      "epoch": 0.6520132269986384,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004962923876236092,
      "loss": 5.2649,
      "mean_token_accuracy": 0.17731465846300126,
      "num_tokens": 14544929.0,
      "step": 8380
    },
    {
      "entropy": 5.623966121673584,
      "epoch": 0.652402256370356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004962873752439035,
      "loss": 5.4207,
      "mean_token_accuracy": 0.16555396616458892,
      "num_tokens": 14553883.0,
      "step": 8385
    },
    {
      "entropy": 5.685714960098267,
      "epoch": 0.6527912857420736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962823595065141,
      "loss": 5.4371,
      "mean_token_accuracy": 0.17250010073184968,
      "num_tokens": 14563085.0,
      "step": 8390
    },
    {
      "entropy": 5.634870195388794,
      "epoch": 0.653180315113791,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962773404115172,
      "loss": 5.3234,
      "mean_token_accuracy": 0.17301514893770217,
      "num_tokens": 14571444.0,
      "step": 8395
    },
    {
      "entropy": 5.61317024230957,
      "epoch": 0.6535693444855086,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004962723179589886,
      "loss": 5.402,
      "mean_token_accuracy": 0.17384381145238875,
      "num_tokens": 14580039.0,
      "step": 8400
    },
    {
      "entropy": 5.569257354736328,
      "epoch": 0.6539583738572262,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000496267292149005,
      "loss": 5.264,
      "mean_token_accuracy": 0.18640352338552474,
      "num_tokens": 14588322.0,
      "step": 8405
    },
    {
      "entropy": 5.545172119140625,
      "epoch": 0.6543474032289438,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004962622629816423,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1789401650428772,
      "num_tokens": 14597503.0,
      "step": 8410
    },
    {
      "entropy": 5.587301301956177,
      "epoch": 0.6547364326006614,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000496257230456977,
      "loss": 5.3655,
      "mean_token_accuracy": 0.17540100067853928,
      "num_tokens": 14605804.0,
      "step": 8415
    },
    {
      "entropy": 5.4942773342132565,
      "epoch": 0.6551254619723789,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004962521945750855,
      "loss": 5.2975,
      "mean_token_accuracy": 0.18086225539445877,
      "num_tokens": 14614064.0,
      "step": 8420
    },
    {
      "entropy": 5.6478150367736815,
      "epoch": 0.6555144913440965,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496247155336044,
      "loss": 5.2648,
      "mean_token_accuracy": 0.18014039248228073,
      "num_tokens": 14622231.0,
      "step": 8425
    },
    {
      "entropy": 5.552904510498047,
      "epoch": 0.655903520715814,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962421127399291,
      "loss": 5.4039,
      "mean_token_accuracy": 0.1692470505833626,
      "num_tokens": 14631244.0,
      "step": 8430
    },
    {
      "entropy": 5.524405384063721,
      "epoch": 0.6562925500875316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962370667868172,
      "loss": 5.2641,
      "mean_token_accuracy": 0.17578798532485962,
      "num_tokens": 14639575.0,
      "step": 8435
    },
    {
      "entropy": 5.594518995285034,
      "epoch": 0.6566815794592492,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962320174767851,
      "loss": 5.3199,
      "mean_token_accuracy": 0.17032575458288193,
      "num_tokens": 14648810.0,
      "step": 8440
    },
    {
      "entropy": 5.552857780456543,
      "epoch": 0.6570706088309667,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496226964809909,
      "loss": 5.1853,
      "mean_token_accuracy": 0.18572386652231215,
      "num_tokens": 14657783.0,
      "step": 8445
    },
    {
      "entropy": 5.627866172790528,
      "epoch": 0.6574596382026843,
      "grad_norm": 1.0,
      "learning_rate": 0.000496221908786266,
      "loss": 5.366,
      "mean_token_accuracy": 0.1744003027677536,
      "num_tokens": 14666872.0,
      "step": 8450
    },
    {
      "entropy": 5.640341329574585,
      "epoch": 0.6578486675744019,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962168494059327,
      "loss": 5.4804,
      "mean_token_accuracy": 0.1615436539053917,
      "num_tokens": 14675762.0,
      "step": 8455
    },
    {
      "entropy": 5.52385983467102,
      "epoch": 0.6582376969461194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004962117866689857,
      "loss": 5.1544,
      "mean_token_accuracy": 0.18567788302898408,
      "num_tokens": 14683928.0,
      "step": 8460
    },
    {
      "entropy": 5.588550329208374,
      "epoch": 0.658626726317837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000496206720575502,
      "loss": 5.4038,
      "mean_token_accuracy": 0.17150904834270478,
      "num_tokens": 14692411.0,
      "step": 8465
    },
    {
      "entropy": 5.625606918334961,
      "epoch": 0.6590157556895546,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004962016511255584,
      "loss": 5.3984,
      "mean_token_accuracy": 0.16342231035232543,
      "num_tokens": 14701002.0,
      "step": 8470
    },
    {
      "entropy": 5.719677305221557,
      "epoch": 0.6594047850612721,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961965783192318,
      "loss": 5.4499,
      "mean_token_accuracy": 0.1649780184030533,
      "num_tokens": 14710172.0,
      "step": 8475
    },
    {
      "entropy": 5.5820893287658695,
      "epoch": 0.6597938144329897,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961915021565992,
      "loss": 5.372,
      "mean_token_accuracy": 0.174167899787426,
      "num_tokens": 14719446.0,
      "step": 8480
    },
    {
      "entropy": 5.643054580688476,
      "epoch": 0.6601828438047073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961864226377377,
      "loss": 5.3936,
      "mean_token_accuracy": 0.17270140647888182,
      "num_tokens": 14728193.0,
      "step": 8485
    },
    {
      "entropy": 5.587178659439087,
      "epoch": 0.6605718731764249,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961813397627241,
      "loss": 5.2796,
      "mean_token_accuracy": 0.1707806646823883,
      "num_tokens": 14736930.0,
      "step": 8490
    },
    {
      "entropy": 5.566015577316284,
      "epoch": 0.6609609025481424,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961762535316358,
      "loss": 5.2835,
      "mean_token_accuracy": 0.18451686799526215,
      "num_tokens": 14745995.0,
      "step": 8495
    },
    {
      "entropy": 5.511094856262207,
      "epoch": 0.6613499319198599,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961711639445499,
      "loss": 5.3234,
      "mean_token_accuracy": 0.16954112946987152,
      "num_tokens": 14754827.0,
      "step": 8500
    },
    {
      "entropy": 5.64048957824707,
      "epoch": 0.6617389612915775,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004961660710015435,
      "loss": 5.3943,
      "mean_token_accuracy": 0.17115921825170516,
      "num_tokens": 14763594.0,
      "step": 8505
    },
    {
      "entropy": 5.57706847190857,
      "epoch": 0.6621279906632951,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004961609747026943,
      "loss": 5.2597,
      "mean_token_accuracy": 0.18465231359004974,
      "num_tokens": 14772366.0,
      "step": 8510
    },
    {
      "entropy": 5.568675518035889,
      "epoch": 0.6625170200350127,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961558750480791,
      "loss": 5.3195,
      "mean_token_accuracy": 0.17469608932733535,
      "num_tokens": 14780738.0,
      "step": 8515
    },
    {
      "entropy": 5.560606145858765,
      "epoch": 0.6629060494067303,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004961507720377756,
      "loss": 5.3064,
      "mean_token_accuracy": 0.17217658460140228,
      "num_tokens": 14788982.0,
      "step": 8520
    },
    {
      "entropy": 5.572122621536255,
      "epoch": 0.6632950787784477,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004961456656718611,
      "loss": 5.2772,
      "mean_token_accuracy": 0.18133418411016464,
      "num_tokens": 14797832.0,
      "step": 8525
    },
    {
      "entropy": 5.604490184783936,
      "epoch": 0.6636841081501653,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496140555950413,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18062032610177994,
      "num_tokens": 14806391.0,
      "step": 8530
    },
    {
      "entropy": 5.632673883438111,
      "epoch": 0.6640731375218829,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961354428735091,
      "loss": 5.3655,
      "mean_token_accuracy": 0.16731194257736207,
      "num_tokens": 14814920.0,
      "step": 8535
    },
    {
      "entropy": 5.536898851394653,
      "epoch": 0.6644621668936005,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961303264412267,
      "loss": 5.2996,
      "mean_token_accuracy": 0.1719740927219391,
      "num_tokens": 14823541.0,
      "step": 8540
    },
    {
      "entropy": 5.521430206298828,
      "epoch": 0.6648511962653181,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961252066536437,
      "loss": 5.2532,
      "mean_token_accuracy": 0.17664670646190644,
      "num_tokens": 14832128.0,
      "step": 8545
    },
    {
      "entropy": 5.618033218383789,
      "epoch": 0.6652402256370356,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961200835108375,
      "loss": 5.4163,
      "mean_token_accuracy": 0.1642109453678131,
      "num_tokens": 14841434.0,
      "step": 8550
    },
    {
      "entropy": 5.580253076553345,
      "epoch": 0.6656292550087531,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496114957012886,
      "loss": 5.3437,
      "mean_token_accuracy": 0.17707033604383468,
      "num_tokens": 14850255.0,
      "step": 8555
    },
    {
      "entropy": 5.596495056152344,
      "epoch": 0.6660182843804707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961098271598671,
      "loss": 5.4239,
      "mean_token_accuracy": 0.16565164923667908,
      "num_tokens": 14858924.0,
      "step": 8560
    },
    {
      "entropy": 5.557281541824341,
      "epoch": 0.6664073137521883,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004961046939518585,
      "loss": 5.3626,
      "mean_token_accuracy": 0.1835842475295067,
      "num_tokens": 14867426.0,
      "step": 8565
    },
    {
      "entropy": 5.608879566192627,
      "epoch": 0.6667963431239059,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960995573889379,
      "loss": 5.3233,
      "mean_token_accuracy": 0.17866379916667938,
      "num_tokens": 14876218.0,
      "step": 8570
    },
    {
      "entropy": 5.628923988342285,
      "epoch": 0.6671853724956234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960944174711836,
      "loss": 5.3501,
      "mean_token_accuracy": 0.17644556611776352,
      "num_tokens": 14883995.0,
      "step": 8575
    },
    {
      "entropy": 5.531912612915039,
      "epoch": 0.667574401867341,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960892741986734,
      "loss": 5.3415,
      "mean_token_accuracy": 0.17149577885866166,
      "num_tokens": 14892659.0,
      "step": 8580
    },
    {
      "entropy": 5.603481960296631,
      "epoch": 0.6679634312390585,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004960841275714853,
      "loss": 5.3565,
      "mean_token_accuracy": 0.17346323430538177,
      "num_tokens": 14901518.0,
      "step": 8585
    },
    {
      "entropy": 5.588987255096436,
      "epoch": 0.6683524606107761,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004960789775896975,
      "loss": 5.3539,
      "mean_token_accuracy": 0.17684564739465714,
      "num_tokens": 14910171.0,
      "step": 8590
    },
    {
      "entropy": 5.598988437652588,
      "epoch": 0.6687414899824937,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960738242533881,
      "loss": 5.3309,
      "mean_token_accuracy": 0.17272143065929413,
      "num_tokens": 14918651.0,
      "step": 8595
    },
    {
      "entropy": 5.54523344039917,
      "epoch": 0.6691305193542112,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960686675626353,
      "loss": 5.3431,
      "mean_token_accuracy": 0.17127756029367447,
      "num_tokens": 14926951.0,
      "step": 8600
    },
    {
      "entropy": 5.638353586196899,
      "epoch": 0.6695195487259288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004960635075175173,
      "loss": 5.4229,
      "mean_token_accuracy": 0.17560173869132994,
      "num_tokens": 14935528.0,
      "step": 8605
    },
    {
      "entropy": 5.578430795669556,
      "epoch": 0.6699085780976464,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960583441181124,
      "loss": 5.2918,
      "mean_token_accuracy": 0.1793479025363922,
      "num_tokens": 14944117.0,
      "step": 8610
    },
    {
      "entropy": 5.4936277866363525,
      "epoch": 0.670297607469364,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960531773644992,
      "loss": 5.281,
      "mean_token_accuracy": 0.17999966442584991,
      "num_tokens": 14952957.0,
      "step": 8615
    },
    {
      "entropy": 5.590290164947509,
      "epoch": 0.6706866368410815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960480072567557,
      "loss": 5.3397,
      "mean_token_accuracy": 0.17372976392507553,
      "num_tokens": 14960942.0,
      "step": 8620
    },
    {
      "entropy": 5.689679288864136,
      "epoch": 0.671075666212799,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004960428337949604,
      "loss": 5.4342,
      "mean_token_accuracy": 0.16590507626533507,
      "num_tokens": 14970579.0,
      "step": 8625
    },
    {
      "entropy": 5.559392261505127,
      "epoch": 0.6714646955845166,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960376569791921,
      "loss": 5.2691,
      "mean_token_accuracy": 0.1740999326109886,
      "num_tokens": 14979353.0,
      "step": 8630
    },
    {
      "entropy": 5.622280597686768,
      "epoch": 0.6718537249562342,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004960324768095291,
      "loss": 5.4223,
      "mean_token_accuracy": 0.1784879356622696,
      "num_tokens": 14988435.0,
      "step": 8635
    },
    {
      "entropy": 5.572164249420166,
      "epoch": 0.6722427543279518,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004960272932860502,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1725583553314209,
      "num_tokens": 14996834.0,
      "step": 8640
    },
    {
      "entropy": 5.635668087005615,
      "epoch": 0.6726317836996694,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960221064088338,
      "loss": 5.4251,
      "mean_token_accuracy": 0.1681579142808914,
      "num_tokens": 15004988.0,
      "step": 8645
    },
    {
      "entropy": 5.679348278045654,
      "epoch": 0.6730208130713868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004960169161779588,
      "loss": 5.364,
      "mean_token_accuracy": 0.1767076998949051,
      "num_tokens": 15013482.0,
      "step": 8650
    },
    {
      "entropy": 5.5645733833312985,
      "epoch": 0.6734098424431044,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960117225935038,
      "loss": 5.256,
      "mean_token_accuracy": 0.1800806149840355,
      "num_tokens": 15021506.0,
      "step": 8655
    },
    {
      "entropy": 5.573781538009643,
      "epoch": 0.673798871814822,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004960065256555477,
      "loss": 5.3037,
      "mean_token_accuracy": 0.18268021196126938,
      "num_tokens": 15029880.0,
      "step": 8660
    },
    {
      "entropy": 5.5966973304748535,
      "epoch": 0.6741879011865396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004960013253641695,
      "loss": 5.3186,
      "mean_token_accuracy": 0.171515029668808,
      "num_tokens": 15038080.0,
      "step": 8665
    },
    {
      "entropy": 5.5642539978027346,
      "epoch": 0.6745769305582572,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959961217194477,
      "loss": 5.3144,
      "mean_token_accuracy": 0.17760654240846635,
      "num_tokens": 15046779.0,
      "step": 8670
    },
    {
      "entropy": 5.587439060211182,
      "epoch": 0.6749659599299748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959909147214615,
      "loss": 5.3541,
      "mean_token_accuracy": 0.17869468182325363,
      "num_tokens": 15055557.0,
      "step": 8675
    },
    {
      "entropy": 5.555358219146728,
      "epoch": 0.6753549893016922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004959857043702902,
      "loss": 5.2256,
      "mean_token_accuracy": 0.17852253168821336,
      "num_tokens": 15063054.0,
      "step": 8680
    },
    {
      "entropy": 5.546805381774902,
      "epoch": 0.6757440186734098,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959804906660124,
      "loss": 5.3019,
      "mean_token_accuracy": 0.17831941694021225,
      "num_tokens": 15071510.0,
      "step": 8685
    },
    {
      "entropy": 5.582370567321777,
      "epoch": 0.6761330480451274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959752736087073,
      "loss": 5.3322,
      "mean_token_accuracy": 0.1665417283773422,
      "num_tokens": 15080524.0,
      "step": 8690
    },
    {
      "entropy": 5.637124013900757,
      "epoch": 0.676522077416845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959700531984543,
      "loss": 5.3323,
      "mean_token_accuracy": 0.17430973425507545,
      "num_tokens": 15089689.0,
      "step": 8695
    },
    {
      "entropy": 5.535669374465942,
      "epoch": 0.6769111067885626,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004959648294353324,
      "loss": 5.2417,
      "mean_token_accuracy": 0.18446339964866637,
      "num_tokens": 15097970.0,
      "step": 8700
    },
    {
      "entropy": 5.506096458435058,
      "epoch": 0.6773001361602801,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004959596023194208,
      "loss": 5.3528,
      "mean_token_accuracy": 0.17490625977516175,
      "num_tokens": 15107786.0,
      "step": 8705
    },
    {
      "entropy": 5.624168634414673,
      "epoch": 0.6776891655319977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495954371850799,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1771204650402069,
      "num_tokens": 15116370.0,
      "step": 8710
    },
    {
      "entropy": 5.643984079360962,
      "epoch": 0.6780781949037152,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004959491380295463,
      "loss": 5.345,
      "mean_token_accuracy": 0.175424787402153,
      "num_tokens": 15124961.0,
      "step": 8715
    },
    {
      "entropy": 5.587209177017212,
      "epoch": 0.6784672242754328,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959439008557422,
      "loss": 5.373,
      "mean_token_accuracy": 0.17338978201150895,
      "num_tokens": 15134230.0,
      "step": 8720
    },
    {
      "entropy": 5.556934022903443,
      "epoch": 0.6788562536471504,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000495938660329466,
      "loss": 5.3099,
      "mean_token_accuracy": 0.1693642869591713,
      "num_tokens": 15143585.0,
      "step": 8725
    },
    {
      "entropy": 5.572135066986084,
      "epoch": 0.6792452830188679,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004959334164507973,
      "loss": 5.2502,
      "mean_token_accuracy": 0.1736395001411438,
      "num_tokens": 15151834.0,
      "step": 8730
    },
    {
      "entropy": 5.513864994049072,
      "epoch": 0.6796343123905855,
      "grad_norm": 1.0,
      "learning_rate": 0.0004959281692198155,
      "loss": 5.3073,
      "mean_token_accuracy": 0.17804726213216782,
      "num_tokens": 15160630.0,
      "step": 8735
    },
    {
      "entropy": 5.5666755676269535,
      "epoch": 0.6800233417623031,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004959229186366007,
      "loss": 5.2752,
      "mean_token_accuracy": 0.17743316441774368,
      "num_tokens": 15168999.0,
      "step": 8740
    },
    {
      "entropy": 5.586798000335693,
      "epoch": 0.6804123711340206,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495917664701232,
      "loss": 5.3409,
      "mean_token_accuracy": 0.18191014379262924,
      "num_tokens": 15177507.0,
      "step": 8745
    },
    {
      "entropy": 5.586767101287842,
      "epoch": 0.6808014005057382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959124074137894,
      "loss": 5.3666,
      "mean_token_accuracy": 0.17173490673303604,
      "num_tokens": 15186843.0,
      "step": 8750
    },
    {
      "entropy": 5.546265268325806,
      "epoch": 0.6811904298774557,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004959071467743526,
      "loss": 5.2778,
      "mean_token_accuracy": 0.1792579934000969,
      "num_tokens": 15195081.0,
      "step": 8755
    },
    {
      "entropy": 5.601288747787476,
      "epoch": 0.6815794592491733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004959018827830016,
      "loss": 5.3662,
      "mean_token_accuracy": 0.1787705212831497,
      "num_tokens": 15203538.0,
      "step": 8760
    },
    {
      "entropy": 5.646421003341675,
      "epoch": 0.6819684886208909,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495896615439816,
      "loss": 5.3994,
      "mean_token_accuracy": 0.17273479998111724,
      "num_tokens": 15211845.0,
      "step": 8765
    },
    {
      "entropy": 5.62197904586792,
      "epoch": 0.6823575179926085,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958913447448759,
      "loss": 5.3891,
      "mean_token_accuracy": 0.1769599959254265,
      "num_tokens": 15220128.0,
      "step": 8770
    },
    {
      "entropy": 5.64114031791687,
      "epoch": 0.682746547364326,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958860706982613,
      "loss": 5.3817,
      "mean_token_accuracy": 0.17129921019077302,
      "num_tokens": 15228741.0,
      "step": 8775
    },
    {
      "entropy": 5.6234687805175785,
      "epoch": 0.6831355767360435,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000495880793300052,
      "loss": 5.3603,
      "mean_token_accuracy": 0.170152148604393,
      "num_tokens": 15237674.0,
      "step": 8780
    },
    {
      "entropy": 5.567347431182862,
      "epoch": 0.6835246061077611,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004958755125503284,
      "loss": 5.2565,
      "mean_token_accuracy": 0.18037977069616318,
      "num_tokens": 15245727.0,
      "step": 8785
    },
    {
      "entropy": 5.617737531661987,
      "epoch": 0.6839136354794787,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004958702284491703,
      "loss": 5.3467,
      "mean_token_accuracy": 0.17623167634010314,
      "num_tokens": 15254511.0,
      "step": 8790
    },
    {
      "entropy": 5.614225721359253,
      "epoch": 0.6843026648511963,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958649409966581,
      "loss": 5.3172,
      "mean_token_accuracy": 0.17816694229841232,
      "num_tokens": 15263500.0,
      "step": 8795
    },
    {
      "entropy": 5.543223428726196,
      "epoch": 0.6846916942229139,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000495859650192872,
      "loss": 5.3265,
      "mean_token_accuracy": 0.17604704052209855,
      "num_tokens": 15271960.0,
      "step": 8800
    },
    {
      "entropy": 5.645513439178467,
      "epoch": 0.6850807235946313,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958543560378922,
      "loss": 5.5038,
      "mean_token_accuracy": 0.16777303814888,
      "num_tokens": 15280576.0,
      "step": 8805
    },
    {
      "entropy": 5.63763108253479,
      "epoch": 0.6854697529663489,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004958490585317991,
      "loss": 5.3633,
      "mean_token_accuracy": 0.17303601950407027,
      "num_tokens": 15288674.0,
      "step": 8810
    },
    {
      "entropy": 5.5421442031860355,
      "epoch": 0.6858587823380665,
      "grad_norm": 1.0625,
      "learning_rate": 0.000495843757674673,
      "loss": 5.2768,
      "mean_token_accuracy": 0.18115932494401932,
      "num_tokens": 15296881.0,
      "step": 8815
    },
    {
      "entropy": 5.580213451385498,
      "epoch": 0.6862478117097841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958384534665945,
      "loss": 5.3204,
      "mean_token_accuracy": 0.17935770750045776,
      "num_tokens": 15305379.0,
      "step": 8820
    },
    {
      "entropy": 5.6232232570648195,
      "epoch": 0.6866368410815017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958331459076438,
      "loss": 5.4179,
      "mean_token_accuracy": 0.16529747992753982,
      "num_tokens": 15314572.0,
      "step": 8825
    },
    {
      "entropy": 5.618761348724365,
      "epoch": 0.6870258704532192,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958278349979018,
      "loss": 5.3705,
      "mean_token_accuracy": 0.1721414342522621,
      "num_tokens": 15323407.0,
      "step": 8830
    },
    {
      "entropy": 5.559399747848511,
      "epoch": 0.6874148998249368,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958225207374488,
      "loss": 5.2481,
      "mean_token_accuracy": 0.17530919909477233,
      "num_tokens": 15332208.0,
      "step": 8835
    },
    {
      "entropy": 5.6014885902404785,
      "epoch": 0.6878039291966543,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958172031263655,
      "loss": 5.3014,
      "mean_token_accuracy": 0.175260029733181,
      "num_tokens": 15340972.0,
      "step": 8840
    },
    {
      "entropy": 5.557904386520386,
      "epoch": 0.6881929585683719,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958118821647326,
      "loss": 5.2855,
      "mean_token_accuracy": 0.18032519072294234,
      "num_tokens": 15349991.0,
      "step": 8845
    },
    {
      "entropy": 5.501622247695923,
      "epoch": 0.6885819879400895,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004958065578526308,
      "loss": 5.239,
      "mean_token_accuracy": 0.18351080715656282,
      "num_tokens": 15358194.0,
      "step": 8850
    },
    {
      "entropy": 5.605360126495361,
      "epoch": 0.688971017311807,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000495801230190141,
      "loss": 5.285,
      "mean_token_accuracy": 0.1786123111844063,
      "num_tokens": 15366954.0,
      "step": 8855
    },
    {
      "entropy": 5.641331720352173,
      "epoch": 0.6893600466835246,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957958991773441,
      "loss": 5.4584,
      "mean_token_accuracy": 0.16349294632673264,
      "num_tokens": 15376107.0,
      "step": 8860
    },
    {
      "entropy": 5.544419240951538,
      "epoch": 0.6897490760552422,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957905648143206,
      "loss": 5.3301,
      "mean_token_accuracy": 0.1729828327894211,
      "num_tokens": 15384517.0,
      "step": 8865
    },
    {
      "entropy": 5.690195560455322,
      "epoch": 0.6901381054269597,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957852271011519,
      "loss": 5.4643,
      "mean_token_accuracy": 0.16566426306962967,
      "num_tokens": 15393544.0,
      "step": 8870
    },
    {
      "entropy": 5.606956481933594,
      "epoch": 0.6905271347986773,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004957798860379187,
      "loss": 5.3012,
      "mean_token_accuracy": 0.17959974557161332,
      "num_tokens": 15402165.0,
      "step": 8875
    },
    {
      "entropy": 5.5568019390106205,
      "epoch": 0.6909161641703949,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957745416247022,
      "loss": 5.3473,
      "mean_token_accuracy": 0.17562471181154252,
      "num_tokens": 15411430.0,
      "step": 8880
    },
    {
      "entropy": 5.581215810775757,
      "epoch": 0.6913051935421124,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957691938615833,
      "loss": 5.2477,
      "mean_token_accuracy": 0.1813277468085289,
      "num_tokens": 15420346.0,
      "step": 8885
    },
    {
      "entropy": 5.449936437606811,
      "epoch": 0.69169422291383,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957638427486434,
      "loss": 5.1729,
      "mean_token_accuracy": 0.1870804876089096,
      "num_tokens": 15428747.0,
      "step": 8890
    },
    {
      "entropy": 5.425696229934692,
      "epoch": 0.6920832522855476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957584882859636,
      "loss": 5.198,
      "mean_token_accuracy": 0.1842238187789917,
      "num_tokens": 15437110.0,
      "step": 8895
    },
    {
      "entropy": 5.560818386077881,
      "epoch": 0.6924722816572652,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957531304736251,
      "loss": 5.3023,
      "mean_token_accuracy": 0.17740701735019684,
      "num_tokens": 15446524.0,
      "step": 8900
    },
    {
      "entropy": 5.557398748397827,
      "epoch": 0.6928613110289827,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957477693117091,
      "loss": 5.367,
      "mean_token_accuracy": 0.17398437559604646,
      "num_tokens": 15455372.0,
      "step": 8905
    },
    {
      "entropy": 5.5803440570831295,
      "epoch": 0.6932503404007002,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004957424048002971,
      "loss": 5.2479,
      "mean_token_accuracy": 0.17678392976522445,
      "num_tokens": 15463060.0,
      "step": 8910
    },
    {
      "entropy": 5.549173164367676,
      "epoch": 0.6936393697724178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004957370369394706,
      "loss": 5.3541,
      "mean_token_accuracy": 0.17353209257125854,
      "num_tokens": 15471836.0,
      "step": 8915
    },
    {
      "entropy": 5.622178792953491,
      "epoch": 0.6940283991441354,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004957316657293107,
      "loss": 5.3608,
      "mean_token_accuracy": 0.17495280802249907,
      "num_tokens": 15479819.0,
      "step": 8920
    },
    {
      "entropy": 5.546295309066773,
      "epoch": 0.694417428515853,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004957262911698992,
      "loss": 5.3473,
      "mean_token_accuracy": 0.1786705568432808,
      "num_tokens": 15488562.0,
      "step": 8925
    },
    {
      "entropy": 5.595986986160279,
      "epoch": 0.6948064578875706,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957209132613175,
      "loss": 5.4397,
      "mean_token_accuracy": 0.1706948459148407,
      "num_tokens": 15497822.0,
      "step": 8930
    },
    {
      "entropy": 5.676778173446655,
      "epoch": 0.695195487259288,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957155320036473,
      "loss": 5.3754,
      "mean_token_accuracy": 0.17063734531402588,
      "num_tokens": 15506913.0,
      "step": 8935
    },
    {
      "entropy": 5.602615642547607,
      "epoch": 0.6955845166310056,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957101473969702,
      "loss": 5.305,
      "mean_token_accuracy": 0.1809922933578491,
      "num_tokens": 15516053.0,
      "step": 8940
    },
    {
      "entropy": 5.616203308105469,
      "epoch": 0.6959735460027232,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495704759441368,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1677024707198143,
      "num_tokens": 15524174.0,
      "step": 8945
    },
    {
      "entropy": 5.590238857269287,
      "epoch": 0.6963625753744408,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004956993681369221,
      "loss": 5.2836,
      "mean_token_accuracy": 0.17359905689954758,
      "num_tokens": 15533652.0,
      "step": 8950
    },
    {
      "entropy": 5.58105239868164,
      "epoch": 0.6967516047461584,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956939734837148,
      "loss": 5.2959,
      "mean_token_accuracy": 0.1745977059006691,
      "num_tokens": 15541644.0,
      "step": 8955
    },
    {
      "entropy": 5.489080953598022,
      "epoch": 0.6971406341178759,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956885754818277,
      "loss": 5.2744,
      "mean_token_accuracy": 0.18329184353351594,
      "num_tokens": 15550298.0,
      "step": 8960
    },
    {
      "entropy": 5.608300447463989,
      "epoch": 0.6975296634895934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004956831741313427,
      "loss": 5.3513,
      "mean_token_accuracy": 0.17609672695398332,
      "num_tokens": 15559388.0,
      "step": 8965
    },
    {
      "entropy": 5.49210147857666,
      "epoch": 0.697918692861311,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004956777694323418,
      "loss": 5.17,
      "mean_token_accuracy": 0.18036870807409286,
      "num_tokens": 15567543.0,
      "step": 8970
    },
    {
      "entropy": 5.508568477630615,
      "epoch": 0.6983077222330286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000495672361384907,
      "loss": 5.2929,
      "mean_token_accuracy": 0.18176358789205552,
      "num_tokens": 15575365.0,
      "step": 8975
    },
    {
      "entropy": 5.6101562023162845,
      "epoch": 0.6986967516047462,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004956669499891202,
      "loss": 5.3468,
      "mean_token_accuracy": 0.1734916463494301,
      "num_tokens": 15584295.0,
      "step": 8980
    },
    {
      "entropy": 5.615858125686645,
      "epoch": 0.6990857809764637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004956615352450639,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18506000936031342,
      "num_tokens": 15592906.0,
      "step": 8985
    },
    {
      "entropy": 5.513598918914795,
      "epoch": 0.6994748103481813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956561171528198,
      "loss": 5.2893,
      "mean_token_accuracy": 0.17286851853132248,
      "num_tokens": 15601938.0,
      "step": 8990
    },
    {
      "entropy": 5.522141981124878,
      "epoch": 0.6998638397198989,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956506957124704,
      "loss": 5.2907,
      "mean_token_accuracy": 0.18730715066194534,
      "num_tokens": 15610173.0,
      "step": 8995
    },
    {
      "entropy": 5.591469383239746,
      "epoch": 0.7002528690916164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000495645270924098,
      "loss": 5.3674,
      "mean_token_accuracy": 0.17109771966934204,
      "num_tokens": 15618896.0,
      "step": 9000
    },
    {
      "epoch": 0.7002528690916164,
      "eval_entropy": 5.334863831241964,
      "eval_loss": 5.353067398071289,
      "eval_mean_token_accuracy": 0.1837482512585238,
      "eval_num_tokens": 15618896.0,
      "eval_runtime": 21.7974,
      "eval_samples_per_second": 1906.557,
      "eval_steps_per_second": 238.331,
      "step": 9000
    },
    {
      "entropy": 5.517790937423706,
      "epoch": 0.700641898463334,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004956398427877847,
      "loss": 5.2849,
      "mean_token_accuracy": 0.18094601929187776,
      "num_tokens": 15627890.0,
      "step": 9005
    },
    {
      "entropy": 5.547005701065063,
      "epoch": 0.7010309278350515,
      "grad_norm": 1.046875,
      "learning_rate": 0.000495634411303613,
      "loss": 5.2258,
      "mean_token_accuracy": 0.18243036270141602,
      "num_tokens": 15636290.0,
      "step": 9010
    },
    {
      "entropy": 5.5310866355896,
      "epoch": 0.7014199572067691,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004956289764716654,
      "loss": 5.2786,
      "mean_token_accuracy": 0.1766011744737625,
      "num_tokens": 15644777.0,
      "step": 9015
    },
    {
      "entropy": 5.514647483825684,
      "epoch": 0.7018089865784867,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956235382920241,
      "loss": 5.3011,
      "mean_token_accuracy": 0.178700989484787,
      "num_tokens": 15653333.0,
      "step": 9020
    },
    {
      "entropy": 5.594124031066895,
      "epoch": 0.7021980159502043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956180967647717,
      "loss": 5.3055,
      "mean_token_accuracy": 0.17473521530628205,
      "num_tokens": 15662140.0,
      "step": 9025
    },
    {
      "entropy": 5.62333984375,
      "epoch": 0.7025870453219218,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004956126518899911,
      "loss": 5.4156,
      "mean_token_accuracy": 0.18038762211799622,
      "num_tokens": 15671387.0,
      "step": 9030
    },
    {
      "entropy": 5.5569250106811525,
      "epoch": 0.7029760746936393,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956072036677644,
      "loss": 5.3447,
      "mean_token_accuracy": 0.17347424179315568,
      "num_tokens": 15680884.0,
      "step": 9035
    },
    {
      "entropy": 5.601095056533813,
      "epoch": 0.7033651040653569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956017520981746,
      "loss": 5.2769,
      "mean_token_accuracy": 0.17625801861286164,
      "num_tokens": 15690030.0,
      "step": 9040
    },
    {
      "entropy": 5.555998992919922,
      "epoch": 0.7037541334370745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955962971813044,
      "loss": 5.3271,
      "mean_token_accuracy": 0.1773602545261383,
      "num_tokens": 15698260.0,
      "step": 9045
    },
    {
      "entropy": 5.484315586090088,
      "epoch": 0.7041431628087921,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004955908389172364,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1841978535056114,
      "num_tokens": 15706275.0,
      "step": 9050
    },
    {
      "entropy": 5.621068859100342,
      "epoch": 0.7045321921805097,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004955853773060536,
      "loss": 5.4395,
      "mean_token_accuracy": 0.17427613586187363,
      "num_tokens": 15715086.0,
      "step": 9055
    },
    {
      "entropy": 5.54253830909729,
      "epoch": 0.7049212215522271,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955799123478388,
      "loss": 5.2346,
      "mean_token_accuracy": 0.18134862631559373,
      "num_tokens": 15723931.0,
      "step": 9060
    },
    {
      "entropy": 5.513597917556763,
      "epoch": 0.7053102509239447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955744440426748,
      "loss": 5.4092,
      "mean_token_accuracy": 0.17476319074630736,
      "num_tokens": 15732905.0,
      "step": 9065
    },
    {
      "entropy": 5.584516191482544,
      "epoch": 0.7056992802956623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004955689723906448,
      "loss": 5.2782,
      "mean_token_accuracy": 0.1718989536166191,
      "num_tokens": 15742355.0,
      "step": 9070
    },
    {
      "entropy": 5.671843004226685,
      "epoch": 0.7060883096673799,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004955634973918318,
      "loss": 5.348,
      "mean_token_accuracy": 0.1742517203092575,
      "num_tokens": 15750924.0,
      "step": 9075
    },
    {
      "entropy": 5.538636636734009,
      "epoch": 0.7064773390390975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955580190463186,
      "loss": 5.2877,
      "mean_token_accuracy": 0.18234916031360626,
      "num_tokens": 15759677.0,
      "step": 9080
    },
    {
      "entropy": 5.52185959815979,
      "epoch": 0.7068663684108151,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004955525373541886,
      "loss": 5.3023,
      "mean_token_accuracy": 0.1736118823289871,
      "num_tokens": 15767875.0,
      "step": 9085
    },
    {
      "entropy": 5.508476829528808,
      "epoch": 0.7072553977825325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955470523155249,
      "loss": 5.2568,
      "mean_token_accuracy": 0.18183113485574723,
      "num_tokens": 15776194.0,
      "step": 9090
    },
    {
      "entropy": 5.555586290359497,
      "epoch": 0.7076444271542501,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955415639304107,
      "loss": 5.3339,
      "mean_token_accuracy": 0.17915569245815277,
      "num_tokens": 15784774.0,
      "step": 9095
    },
    {
      "entropy": 5.571475601196289,
      "epoch": 0.7080334565259677,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004955360721989292,
      "loss": 5.2312,
      "mean_token_accuracy": 0.18535639494657516,
      "num_tokens": 15793121.0,
      "step": 9100
    },
    {
      "entropy": 5.70520281791687,
      "epoch": 0.7084224858976853,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955305771211641,
      "loss": 5.4456,
      "mean_token_accuracy": 0.17373489290475846,
      "num_tokens": 15801518.0,
      "step": 9105
    },
    {
      "entropy": 5.551429510116577,
      "epoch": 0.7088115152694029,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004955250786971982,
      "loss": 5.2196,
      "mean_token_accuracy": 0.18320250064134597,
      "num_tokens": 15809602.0,
      "step": 9110
    },
    {
      "entropy": 5.551525545120239,
      "epoch": 0.7092005446411204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955195769271154,
      "loss": 5.3978,
      "mean_token_accuracy": 0.17417898178100585,
      "num_tokens": 15818529.0,
      "step": 9115
    },
    {
      "entropy": 5.654647159576416,
      "epoch": 0.709589574012838,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004955140718109991,
      "loss": 5.2735,
      "mean_token_accuracy": 0.1795146495103836,
      "num_tokens": 15827176.0,
      "step": 9120
    },
    {
      "entropy": 5.622618007659912,
      "epoch": 0.7099786033845555,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955085633489326,
      "loss": 5.3541,
      "mean_token_accuracy": 0.1693027675151825,
      "num_tokens": 15837095.0,
      "step": 9125
    },
    {
      "entropy": 5.559453916549683,
      "epoch": 0.7103676327562731,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004955030515409997,
      "loss": 5.3267,
      "mean_token_accuracy": 0.17723278254270552,
      "num_tokens": 15845285.0,
      "step": 9130
    },
    {
      "entropy": 5.569735670089722,
      "epoch": 0.7107566621279907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954975363872838,
      "loss": 5.3331,
      "mean_token_accuracy": 0.18113620728254318,
      "num_tokens": 15853468.0,
      "step": 9135
    },
    {
      "entropy": 5.589026069641113,
      "epoch": 0.7111456914997082,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954920178878689,
      "loss": 5.2715,
      "mean_token_accuracy": 0.1886820301413536,
      "num_tokens": 15861699.0,
      "step": 9140
    },
    {
      "entropy": 5.571455717086792,
      "epoch": 0.7115347208714258,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004954864960428385,
      "loss": 5.3525,
      "mean_token_accuracy": 0.1744161754846573,
      "num_tokens": 15870401.0,
      "step": 9145
    },
    {
      "entropy": 5.542234897613525,
      "epoch": 0.7119237502431434,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004954809708522765,
      "loss": 5.1695,
      "mean_token_accuracy": 0.19075365066528321,
      "num_tokens": 15878214.0,
      "step": 9150
    },
    {
      "entropy": 5.565649938583374,
      "epoch": 0.712312779614861,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954754423162667,
      "loss": 5.3442,
      "mean_token_accuracy": 0.17487907856702806,
      "num_tokens": 15886581.0,
      "step": 9155
    },
    {
      "entropy": 5.639461469650269,
      "epoch": 0.7127018089865785,
      "grad_norm": 0.875,
      "learning_rate": 0.000495469910434893,
      "loss": 5.4492,
      "mean_token_accuracy": 0.1683066964149475,
      "num_tokens": 15895748.0,
      "step": 9160
    },
    {
      "entropy": 5.583930349349975,
      "epoch": 0.713090838358296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954643752082392,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1784547373652458,
      "num_tokens": 15905030.0,
      "step": 9165
    },
    {
      "entropy": 5.546288728713989,
      "epoch": 0.7134798677300136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954588366363896,
      "loss": 5.2954,
      "mean_token_accuracy": 0.17823144495487214,
      "num_tokens": 15913430.0,
      "step": 9170
    },
    {
      "entropy": 5.564022445678711,
      "epoch": 0.7138688971017312,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004954532947194279,
      "loss": 5.1921,
      "mean_token_accuracy": 0.18421580493450165,
      "num_tokens": 15921432.0,
      "step": 9175
    },
    {
      "entropy": 5.593471193313599,
      "epoch": 0.7142579264734488,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004954477494574384,
      "loss": 5.2702,
      "mean_token_accuracy": 0.18055691719055175,
      "num_tokens": 15930159.0,
      "step": 9180
    },
    {
      "entropy": 5.570560455322266,
      "epoch": 0.7146469558451664,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004954422008505052,
      "loss": 5.3586,
      "mean_token_accuracy": 0.17044297009706497,
      "num_tokens": 15939911.0,
      "step": 9185
    },
    {
      "entropy": 5.592472934722901,
      "epoch": 0.7150359852168838,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954366488987125,
      "loss": 5.2855,
      "mean_token_accuracy": 0.17620354592800141,
      "num_tokens": 15948586.0,
      "step": 9190
    },
    {
      "entropy": 5.444886636734009,
      "epoch": 0.7154250145886014,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954310936021444,
      "loss": 5.1423,
      "mean_token_accuracy": 0.18131237626075744,
      "num_tokens": 15956818.0,
      "step": 9195
    },
    {
      "entropy": 5.540569067001343,
      "epoch": 0.715814043960319,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004954255349608853,
      "loss": 5.3682,
      "mean_token_accuracy": 0.16918419450521469,
      "num_tokens": 15966027.0,
      "step": 9200
    },
    {
      "entropy": 5.643780088424682,
      "epoch": 0.7162030733320366,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954199729750198,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17508964836597443,
      "num_tokens": 15975186.0,
      "step": 9205
    },
    {
      "entropy": 5.526380300521851,
      "epoch": 0.7165921027037542,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004954144076446318,
      "loss": 5.2527,
      "mean_token_accuracy": 0.1784766599535942,
      "num_tokens": 15983924.0,
      "step": 9210
    },
    {
      "entropy": 5.552653026580811,
      "epoch": 0.7169811320754716,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954088389698061,
      "loss": 5.2999,
      "mean_token_accuracy": 0.1809309482574463,
      "num_tokens": 15992906.0,
      "step": 9215
    },
    {
      "entropy": 5.587754011154175,
      "epoch": 0.7173701614471892,
      "grad_norm": 0.984375,
      "learning_rate": 0.000495403266950627,
      "loss": 5.3018,
      "mean_token_accuracy": 0.17490899711847305,
      "num_tokens": 16001378.0,
      "step": 9220
    },
    {
      "entropy": 5.538111400604248,
      "epoch": 0.7177591908189068,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004953976915871792,
      "loss": 5.3068,
      "mean_token_accuracy": 0.17638416737318038,
      "num_tokens": 16010704.0,
      "step": 9225
    },
    {
      "entropy": 5.489271974563598,
      "epoch": 0.7181482201906244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953921128795472,
      "loss": 5.2907,
      "mean_token_accuracy": 0.18120024502277374,
      "num_tokens": 16018936.0,
      "step": 9230
    },
    {
      "entropy": 5.587630128860473,
      "epoch": 0.718537249562342,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004953865308278156,
      "loss": 5.19,
      "mean_token_accuracy": 0.17775632292032242,
      "num_tokens": 16027295.0,
      "step": 9235
    },
    {
      "entropy": 5.526306295394898,
      "epoch": 0.7189262789340595,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953809454320693,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1811777248978615,
      "num_tokens": 16036493.0,
      "step": 9240
    },
    {
      "entropy": 5.5192351818084715,
      "epoch": 0.7193153083057771,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953753566923929,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1811526209115982,
      "num_tokens": 16045406.0,
      "step": 9245
    },
    {
      "entropy": 5.559315776824951,
      "epoch": 0.7197043376774946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953697646088712,
      "loss": 5.3211,
      "mean_token_accuracy": 0.17802902311086655,
      "num_tokens": 16054126.0,
      "step": 9250
    },
    {
      "entropy": 5.60638656616211,
      "epoch": 0.7200933670492122,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953641691815891,
      "loss": 5.3362,
      "mean_token_accuracy": 0.170424747467041,
      "num_tokens": 16063921.0,
      "step": 9255
    },
    {
      "entropy": 5.560848045349121,
      "epoch": 0.7204823964209298,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004953585704106315,
      "loss": 5.3469,
      "mean_token_accuracy": 0.17578511089086532,
      "num_tokens": 16072160.0,
      "step": 9260
    },
    {
      "entropy": 5.616014766693115,
      "epoch": 0.7208714257926473,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953529682960832,
      "loss": 5.2744,
      "mean_token_accuracy": 0.17761323899030684,
      "num_tokens": 16080794.0,
      "step": 9265
    },
    {
      "entropy": 5.540022087097168,
      "epoch": 0.7212604551643649,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953473628380295,
      "loss": 5.3094,
      "mean_token_accuracy": 0.18073517382144927,
      "num_tokens": 16089615.0,
      "step": 9270
    },
    {
      "entropy": 5.527288341522217,
      "epoch": 0.7216494845360825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953417540365553,
      "loss": 5.2096,
      "mean_token_accuracy": 0.18059678077697755,
      "num_tokens": 16098363.0,
      "step": 9275
    },
    {
      "entropy": 5.5765989303588865,
      "epoch": 0.7220385139078,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004953361418917455,
      "loss": 5.3158,
      "mean_token_accuracy": 0.17089972347021104,
      "num_tokens": 16106837.0,
      "step": 9280
    },
    {
      "entropy": 5.591169786453247,
      "epoch": 0.7224275432795176,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004953305264036856,
      "loss": 5.2657,
      "mean_token_accuracy": 0.17635613232851027,
      "num_tokens": 16114923.0,
      "step": 9285
    },
    {
      "entropy": 5.5530914783477785,
      "epoch": 0.7228165726512352,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953249075724607,
      "loss": 5.306,
      "mean_token_accuracy": 0.17418773621320724,
      "num_tokens": 16123698.0,
      "step": 9290
    },
    {
      "entropy": 5.58996548652649,
      "epoch": 0.7232056020229527,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953192853981559,
      "loss": 5.2645,
      "mean_token_accuracy": 0.17848737239837648,
      "num_tokens": 16132361.0,
      "step": 9295
    },
    {
      "entropy": 5.63656325340271,
      "epoch": 0.7235946313946703,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953136598808566,
      "loss": 5.3517,
      "mean_token_accuracy": 0.17827641367912292,
      "num_tokens": 16141554.0,
      "step": 9300
    },
    {
      "entropy": 5.569615173339844,
      "epoch": 0.7239836607663879,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953080310206481,
      "loss": 5.2606,
      "mean_token_accuracy": 0.17871276885271073,
      "num_tokens": 16150206.0,
      "step": 9305
    },
    {
      "entropy": 5.556385135650634,
      "epoch": 0.7243726901381055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953023988176162,
      "loss": 5.383,
      "mean_token_accuracy": 0.17167091518640518,
      "num_tokens": 16158995.0,
      "step": 9310
    },
    {
      "entropy": 5.579621839523315,
      "epoch": 0.724761719509823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004952967632718458,
      "loss": 5.3611,
      "mean_token_accuracy": 0.1680089145898819,
      "num_tokens": 16167809.0,
      "step": 9315
    },
    {
      "entropy": 5.725109624862671,
      "epoch": 0.7251507488815405,
      "grad_norm": 1.125,
      "learning_rate": 0.0004952911243834227,
      "loss": 5.4358,
      "mean_token_accuracy": 0.1652451917529106,
      "num_tokens": 16176257.0,
      "step": 9320
    },
    {
      "entropy": 5.625408172607422,
      "epoch": 0.7255397782532581,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952854821524324,
      "loss": 5.3858,
      "mean_token_accuracy": 0.17146064341068268,
      "num_tokens": 16185597.0,
      "step": 9325
    },
    {
      "entropy": 5.643079853057861,
      "epoch": 0.7259288076249757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004952798365789606,
      "loss": 5.4106,
      "mean_token_accuracy": 0.17838159650564195,
      "num_tokens": 16194591.0,
      "step": 9330
    },
    {
      "entropy": 5.683854055404663,
      "epoch": 0.7263178369966933,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952741876630928,
      "loss": 5.3713,
      "mean_token_accuracy": 0.1727147802710533,
      "num_tokens": 16203073.0,
      "step": 9335
    },
    {
      "entropy": 5.543218660354614,
      "epoch": 0.7267068663684109,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952685354049148,
      "loss": 5.3406,
      "mean_token_accuracy": 0.17130525261163712,
      "num_tokens": 16211208.0,
      "step": 9340
    },
    {
      "entropy": 5.565604162216187,
      "epoch": 0.7270958957401283,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004952628798045124,
      "loss": 5.2862,
      "mean_token_accuracy": 0.18018544763326644,
      "num_tokens": 16219687.0,
      "step": 9345
    },
    {
      "entropy": 5.624908447265625,
      "epoch": 0.7274849251118459,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004952572208619714,
      "loss": 5.376,
      "mean_token_accuracy": 0.17842526137828826,
      "num_tokens": 16228214.0,
      "step": 9350
    },
    {
      "entropy": 5.600245380401612,
      "epoch": 0.7278739544835635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952515585773778,
      "loss": 5.4571,
      "mean_token_accuracy": 0.16715698540210724,
      "num_tokens": 16237765.0,
      "step": 9355
    },
    {
      "entropy": 5.585886287689209,
      "epoch": 0.7282629838552811,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004952458929508171,
      "loss": 5.3104,
      "mean_token_accuracy": 0.18905370384454728,
      "num_tokens": 16245339.0,
      "step": 9360
    },
    {
      "entropy": 5.537863397598267,
      "epoch": 0.7286520132269987,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952402239823757,
      "loss": 5.3218,
      "mean_token_accuracy": 0.17315308898687362,
      "num_tokens": 16253966.0,
      "step": 9365
    },
    {
      "entropy": 5.545432424545288,
      "epoch": 0.7290410425987162,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004952345516721393,
      "loss": 5.3002,
      "mean_token_accuracy": 0.17695194780826567,
      "num_tokens": 16262859.0,
      "step": 9370
    },
    {
      "entropy": 5.568144798278809,
      "epoch": 0.7294300719704337,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004952288760201942,
      "loss": 5.3316,
      "mean_token_accuracy": 0.17465603053569795,
      "num_tokens": 16271646.0,
      "step": 9375
    },
    {
      "entropy": 5.563733291625977,
      "epoch": 0.7298191013421513,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952231970266265,
      "loss": 5.3223,
      "mean_token_accuracy": 0.17299434542655945,
      "num_tokens": 16281058.0,
      "step": 9380
    },
    {
      "entropy": 5.645624017715454,
      "epoch": 0.7302081307138689,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952175146915223,
      "loss": 5.2648,
      "mean_token_accuracy": 0.18165217638015746,
      "num_tokens": 16289039.0,
      "step": 9385
    },
    {
      "entropy": 5.457538843154907,
      "epoch": 0.7305971600855865,
      "grad_norm": 1.0,
      "learning_rate": 0.0004952118290149679,
      "loss": 5.1743,
      "mean_token_accuracy": 0.18510949462652207,
      "num_tokens": 16297373.0,
      "step": 9390
    },
    {
      "entropy": 5.541280889511109,
      "epoch": 0.730986189457304,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952061399970492,
      "loss": 5.3314,
      "mean_token_accuracy": 0.17273576706647872,
      "num_tokens": 16306066.0,
      "step": 9395
    },
    {
      "entropy": 5.610443115234375,
      "epoch": 0.7313752188290216,
      "grad_norm": 1.0,
      "learning_rate": 0.000495200447637853,
      "loss": 5.3794,
      "mean_token_accuracy": 0.17505713552236557,
      "num_tokens": 16315282.0,
      "step": 9400
    },
    {
      "entropy": 5.7012941360473635,
      "epoch": 0.7317642482007392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951947519374655,
      "loss": 5.4693,
      "mean_token_accuracy": 0.17104714810848237,
      "num_tokens": 16324830.0,
      "step": 9405
    },
    {
      "entropy": 5.596612167358399,
      "epoch": 0.7321532775724567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004951890528959731,
      "loss": 5.3319,
      "mean_token_accuracy": 0.17324616760015488,
      "num_tokens": 16332990.0,
      "step": 9410
    },
    {
      "entropy": 5.536366128921509,
      "epoch": 0.7325423069441743,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951833505134623,
      "loss": 5.2832,
      "mean_token_accuracy": 0.17535036206245422,
      "num_tokens": 16342115.0,
      "step": 9415
    },
    {
      "entropy": 5.560042858123779,
      "epoch": 0.7329313363158918,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004951776447900196,
      "loss": 5.2877,
      "mean_token_accuracy": 0.17931381464004517,
      "num_tokens": 16350329.0,
      "step": 9420
    },
    {
      "entropy": 5.573030281066894,
      "epoch": 0.7333203656876094,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951719357257317,
      "loss": 5.3897,
      "mean_token_accuracy": 0.17676016688346863,
      "num_tokens": 16358890.0,
      "step": 9425
    },
    {
      "entropy": 5.683777379989624,
      "epoch": 0.733709395059327,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495166223320685,
      "loss": 5.4258,
      "mean_token_accuracy": 0.16870919317007066,
      "num_tokens": 16367338.0,
      "step": 9430
    },
    {
      "entropy": 5.647769021987915,
      "epoch": 0.7340984244310446,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004951605075749662,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1818474978208542,
      "num_tokens": 16375472.0,
      "step": 9435
    },
    {
      "entropy": 5.610141563415527,
      "epoch": 0.7344874538027621,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004951547884886623,
      "loss": 5.3214,
      "mean_token_accuracy": 0.1809833511710167,
      "num_tokens": 16384212.0,
      "step": 9440
    },
    {
      "entropy": 5.5480530738830565,
      "epoch": 0.7348764831744796,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951490660618598,
      "loss": 5.2098,
      "mean_token_accuracy": 0.18420309871435164,
      "num_tokens": 16393447.0,
      "step": 9445
    },
    {
      "entropy": 5.487442255020142,
      "epoch": 0.7352655125461972,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951433402946457,
      "loss": 5.2002,
      "mean_token_accuracy": 0.18515112549066542,
      "num_tokens": 16401876.0,
      "step": 9450
    },
    {
      "entropy": 5.539979934692383,
      "epoch": 0.7356545419179148,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951376111871068,
      "loss": 5.3543,
      "mean_token_accuracy": 0.17673734128475188,
      "num_tokens": 16411020.0,
      "step": 9455
    },
    {
      "entropy": 5.639129781723023,
      "epoch": 0.7360435712896324,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004951318787393299,
      "loss": 5.3708,
      "mean_token_accuracy": 0.17495454996824264,
      "num_tokens": 16420492.0,
      "step": 9460
    },
    {
      "entropy": 5.582475185394287,
      "epoch": 0.73643260066135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951261429514023,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18109755516052245,
      "num_tokens": 16429277.0,
      "step": 9465
    },
    {
      "entropy": 5.522632169723511,
      "epoch": 0.7368216300330674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951204038234106,
      "loss": 5.3087,
      "mean_token_accuracy": 0.17764748632907867,
      "num_tokens": 16438008.0,
      "step": 9470
    },
    {
      "entropy": 5.581943416595459,
      "epoch": 0.737210659404785,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951146613554424,
      "loss": 5.3153,
      "mean_token_accuracy": 0.17503933608531952,
      "num_tokens": 16446791.0,
      "step": 9475
    },
    {
      "entropy": 5.578238868713379,
      "epoch": 0.7375996887765026,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951089155475843,
      "loss": 5.3507,
      "mean_token_accuracy": 0.17409266233444215,
      "num_tokens": 16455277.0,
      "step": 9480
    },
    {
      "entropy": 5.538052940368653,
      "epoch": 0.7379887181482202,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495103166399924,
      "loss": 5.2642,
      "mean_token_accuracy": 0.17929996103048323,
      "num_tokens": 16463310.0,
      "step": 9485
    },
    {
      "entropy": 5.547290706634522,
      "epoch": 0.7383777475199378,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950974139125484,
      "loss": 5.313,
      "mean_token_accuracy": 0.17705847173929215,
      "num_tokens": 16472421.0,
      "step": 9490
    },
    {
      "entropy": 5.575832319259644,
      "epoch": 0.7387667768916554,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004950916580855448,
      "loss": 5.3739,
      "mean_token_accuracy": 0.17194296568632125,
      "num_tokens": 16481814.0,
      "step": 9495
    },
    {
      "entropy": 5.583070898056031,
      "epoch": 0.7391558062633728,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950858989190007,
      "loss": 5.328,
      "mean_token_accuracy": 0.17030642926692963,
      "num_tokens": 16490902.0,
      "step": 9500
    },
    {
      "entropy": 5.6068895816802975,
      "epoch": 0.7395448356350904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950801364130032,
      "loss": 5.2863,
      "mean_token_accuracy": 0.17758143097162246,
      "num_tokens": 16499418.0,
      "step": 9505
    },
    {
      "entropy": 5.531932544708252,
      "epoch": 0.739933865006808,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950743705676401,
      "loss": 5.2203,
      "mean_token_accuracy": 0.18528714179992675,
      "num_tokens": 16508578.0,
      "step": 9510
    },
    {
      "entropy": 5.513550710678101,
      "epoch": 0.7403228943785256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950686013829987,
      "loss": 5.284,
      "mean_token_accuracy": 0.17520827203989028,
      "num_tokens": 16517356.0,
      "step": 9515
    },
    {
      "entropy": 5.5130551815032955,
      "epoch": 0.7407119237502432,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004950628288591665,
      "loss": 5.2898,
      "mean_token_accuracy": 0.17843545973300934,
      "num_tokens": 16525834.0,
      "step": 9520
    },
    {
      "entropy": 5.577107238769531,
      "epoch": 0.7411009531219607,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004950570529962311,
      "loss": 5.2891,
      "mean_token_accuracy": 0.1758483812212944,
      "num_tokens": 16534513.0,
      "step": 9525
    },
    {
      "entropy": 5.60887508392334,
      "epoch": 0.7414899824936783,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950512737942803,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1797830954194069,
      "num_tokens": 16543326.0,
      "step": 9530
    },
    {
      "entropy": 5.548036289215088,
      "epoch": 0.7418790118653958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004950454912534015,
      "loss": 5.3027,
      "mean_token_accuracy": 0.17792366594076156,
      "num_tokens": 16551589.0,
      "step": 9535
    },
    {
      "entropy": 5.582470273971557,
      "epoch": 0.7422680412371134,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004950397053736827,
      "loss": 5.2813,
      "mean_token_accuracy": 0.1794194683432579,
      "num_tokens": 16560261.0,
      "step": 9540
    },
    {
      "entropy": 5.607489442825317,
      "epoch": 0.742657070608831,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950339161552117,
      "loss": 5.269,
      "mean_token_accuracy": 0.1809129610657692,
      "num_tokens": 16568645.0,
      "step": 9545
    },
    {
      "entropy": 5.538193416595459,
      "epoch": 0.7430460999805485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950281235980761,
      "loss": 5.272,
      "mean_token_accuracy": 0.18645498007535935,
      "num_tokens": 16577206.0,
      "step": 9550
    },
    {
      "entropy": 5.541358613967896,
      "epoch": 0.7434351293522661,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495022327702364,
      "loss": 5.2744,
      "mean_token_accuracy": 0.17665404379367827,
      "num_tokens": 16585011.0,
      "step": 9555
    },
    {
      "entropy": 5.591173839569092,
      "epoch": 0.7438241587239837,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004950165284681631,
      "loss": 5.3375,
      "mean_token_accuracy": 0.17531904727220535,
      "num_tokens": 16593574.0,
      "step": 9560
    },
    {
      "entropy": 5.570319128036499,
      "epoch": 0.7442131880957013,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004950107258955618,
      "loss": 5.2482,
      "mean_token_accuracy": 0.18688714802265166,
      "num_tokens": 16603023.0,
      "step": 9565
    },
    {
      "entropy": 5.637556695938111,
      "epoch": 0.7446022174674188,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950049199846479,
      "loss": 5.3453,
      "mean_token_accuracy": 0.17867223173379898,
      "num_tokens": 16611963.0,
      "step": 9570
    },
    {
      "entropy": 5.639278364181519,
      "epoch": 0.7449912468391363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004949991107355095,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18264320790767669,
      "num_tokens": 16619942.0,
      "step": 9575
    },
    {
      "entropy": 5.478429794311523,
      "epoch": 0.7453802762108539,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949932981482347,
      "loss": 5.187,
      "mean_token_accuracy": 0.1897456482052803,
      "num_tokens": 16628261.0,
      "step": 9580
    },
    {
      "entropy": 5.585916566848755,
      "epoch": 0.7457693055825715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004949874822229118,
      "loss": 5.3198,
      "mean_token_accuracy": 0.17921436578035355,
      "num_tokens": 16637290.0,
      "step": 9585
    },
    {
      "entropy": 5.578394842147827,
      "epoch": 0.7461583349542891,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949816629596288,
      "loss": 5.3102,
      "mean_token_accuracy": 0.17700878232717515,
      "num_tokens": 16646147.0,
      "step": 9590
    },
    {
      "entropy": 5.5497089385986325,
      "epoch": 0.7465473643260067,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949758403584746,
      "loss": 5.3844,
      "mean_token_accuracy": 0.16975755989551544,
      "num_tokens": 16655299.0,
      "step": 9595
    },
    {
      "entropy": 5.574647665023804,
      "epoch": 0.7469363936977241,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004949700144195368,
      "loss": 5.2742,
      "mean_token_accuracy": 0.18338656425476074,
      "num_tokens": 16664091.0,
      "step": 9600
    },
    {
      "entropy": 5.692896175384521,
      "epoch": 0.7473254230694417,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949641851429043,
      "loss": 5.3384,
      "mean_token_accuracy": 0.17343031615018845,
      "num_tokens": 16672665.0,
      "step": 9605
    },
    {
      "entropy": 5.559443998336792,
      "epoch": 0.7477144524411593,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949583525286654,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18353851437568663,
      "num_tokens": 16680735.0,
      "step": 9610
    },
    {
      "entropy": 5.589480638504028,
      "epoch": 0.7481034818128769,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949525165769085,
      "loss": 5.3597,
      "mean_token_accuracy": 0.1735190451145172,
      "num_tokens": 16689560.0,
      "step": 9615
    },
    {
      "entropy": 5.60034704208374,
      "epoch": 0.7484925111845945,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949466772877224,
      "loss": 5.3202,
      "mean_token_accuracy": 0.17463106364011766,
      "num_tokens": 16698953.0,
      "step": 9620
    },
    {
      "entropy": 5.597173357009888,
      "epoch": 0.748881540556312,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004949408346611955,
      "loss": 5.3247,
      "mean_token_accuracy": 0.17516023367643357,
      "num_tokens": 16707601.0,
      "step": 9625
    },
    {
      "entropy": 5.54741678237915,
      "epoch": 0.7492705699280295,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949349886974165,
      "loss": 5.3319,
      "mean_token_accuracy": 0.1804911673069,
      "num_tokens": 16716660.0,
      "step": 9630
    },
    {
      "entropy": 5.554853725433349,
      "epoch": 0.7496595992997471,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949291393964741,
      "loss": 5.2625,
      "mean_token_accuracy": 0.1796331986784935,
      "num_tokens": 16725274.0,
      "step": 9635
    },
    {
      "entropy": 5.558338928222656,
      "epoch": 0.7500486286714647,
      "grad_norm": 1.0625,
      "learning_rate": 0.000494923286758457,
      "loss": 5.2656,
      "mean_token_accuracy": 0.17535172253847123,
      "num_tokens": 16733301.0,
      "step": 9640
    },
    {
      "entropy": 5.577684450149536,
      "epoch": 0.7504376580431823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004949174307834541,
      "loss": 5.2874,
      "mean_token_accuracy": 0.17794075906276702,
      "num_tokens": 16742072.0,
      "step": 9645
    },
    {
      "entropy": 5.61627197265625,
      "epoch": 0.7508266874148998,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949115714715543,
      "loss": 5.2957,
      "mean_token_accuracy": 0.17773596048355103,
      "num_tokens": 16751497.0,
      "step": 9650
    },
    {
      "entropy": 5.542064380645752,
      "epoch": 0.7512157167866174,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004949057088228465,
      "loss": 5.2547,
      "mean_token_accuracy": 0.183311328291893,
      "num_tokens": 16760518.0,
      "step": 9655
    },
    {
      "entropy": 5.613554763793945,
      "epoch": 0.751604746158335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948998428374194,
      "loss": 5.3787,
      "mean_token_accuracy": 0.17053319960832597,
      "num_tokens": 16768689.0,
      "step": 9660
    },
    {
      "entropy": 5.582167243957519,
      "epoch": 0.7519937755300525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004948939735153622,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19048200994729997,
      "num_tokens": 16777047.0,
      "step": 9665
    },
    {
      "entropy": 5.49366455078125,
      "epoch": 0.7523828049017701,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948881008567641,
      "loss": 5.1831,
      "mean_token_accuracy": 0.1859971195459366,
      "num_tokens": 16785402.0,
      "step": 9670
    },
    {
      "entropy": 5.553354263305664,
      "epoch": 0.7527718342734876,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948822248617139,
      "loss": 5.2923,
      "mean_token_accuracy": 0.17727821469306945,
      "num_tokens": 16794687.0,
      "step": 9675
    },
    {
      "entropy": 5.621946334838867,
      "epoch": 0.7531608636452052,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000494876345530301,
      "loss": 5.323,
      "mean_token_accuracy": 0.17618732303380966,
      "num_tokens": 16803762.0,
      "step": 9680
    },
    {
      "entropy": 5.583499431610107,
      "epoch": 0.7535498930169228,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004948704628626145,
      "loss": 5.3521,
      "mean_token_accuracy": 0.171107779443264,
      "num_tokens": 16812872.0,
      "step": 9685
    },
    {
      "entropy": 5.652652168273926,
      "epoch": 0.7539389223886404,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004948645768587437,
      "loss": 5.3799,
      "mean_token_accuracy": 0.16893844902515412,
      "num_tokens": 16821270.0,
      "step": 9690
    },
    {
      "entropy": 5.560877323150635,
      "epoch": 0.7543279517603579,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948586875187778,
      "loss": 5.2783,
      "mean_token_accuracy": 0.181769460439682,
      "num_tokens": 16829508.0,
      "step": 9695
    },
    {
      "entropy": 5.560929203033448,
      "epoch": 0.7547169811320755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948527948428064,
      "loss": 5.2748,
      "mean_token_accuracy": 0.17611613422632216,
      "num_tokens": 16838327.0,
      "step": 9700
    },
    {
      "entropy": 5.599258136749268,
      "epoch": 0.755106010503793,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004948468988309187,
      "loss": 5.2464,
      "mean_token_accuracy": 0.18367060720920564,
      "num_tokens": 16846530.0,
      "step": 9705
    },
    {
      "entropy": 5.504384708404541,
      "epoch": 0.7554950398755106,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004948409994832043,
      "loss": 5.1886,
      "mean_token_accuracy": 0.18293901830911635,
      "num_tokens": 16854825.0,
      "step": 9710
    },
    {
      "entropy": 5.591901350021362,
      "epoch": 0.7558840692472282,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004948350967997526,
      "loss": 5.3453,
      "mean_token_accuracy": 0.16848235130310057,
      "num_tokens": 16864096.0,
      "step": 9715
    },
    {
      "entropy": 5.636907529830933,
      "epoch": 0.7562730986189458,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948291907806532,
      "loss": 5.391,
      "mean_token_accuracy": 0.16900479644536973,
      "num_tokens": 16873000.0,
      "step": 9720
    },
    {
      "entropy": 5.680421686172485,
      "epoch": 0.7566621279906633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948232814259957,
      "loss": 5.3942,
      "mean_token_accuracy": 0.17993731200695037,
      "num_tokens": 16881374.0,
      "step": 9725
    },
    {
      "entropy": 5.630921649932861,
      "epoch": 0.7570511573623808,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004948173687358699,
      "loss": 5.3637,
      "mean_token_accuracy": 0.174547378718853,
      "num_tokens": 16889192.0,
      "step": 9730
    },
    {
      "entropy": 5.5700239658355715,
      "epoch": 0.7574401867340984,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004948114527103652,
      "loss": 5.3546,
      "mean_token_accuracy": 0.17323452830314637,
      "num_tokens": 16898538.0,
      "step": 9735
    },
    {
      "entropy": 5.59753451347351,
      "epoch": 0.757829216105816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004948055333495717,
      "loss": 5.343,
      "mean_token_accuracy": 0.1698336720466614,
      "num_tokens": 16907543.0,
      "step": 9740
    },
    {
      "entropy": 5.5732612133026125,
      "epoch": 0.7582182454775336,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004947996106535791,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18186050951480864,
      "num_tokens": 16916140.0,
      "step": 9745
    },
    {
      "entropy": 5.5288232326507565,
      "epoch": 0.7586072748492512,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947936846224773,
      "loss": 5.2556,
      "mean_token_accuracy": 0.18514294773340226,
      "num_tokens": 16925433.0,
      "step": 9750
    },
    {
      "entropy": 5.565367221832275,
      "epoch": 0.7589963042209686,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004947877552563562,
      "loss": 5.2704,
      "mean_token_accuracy": 0.1804751992225647,
      "num_tokens": 16933598.0,
      "step": 9755
    },
    {
      "entropy": 5.6109593391418455,
      "epoch": 0.7593853335926862,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004947818225553056,
      "loss": 5.3761,
      "mean_token_accuracy": 0.1719370514154434,
      "num_tokens": 16942270.0,
      "step": 9760
    },
    {
      "entropy": 5.565868473052978,
      "epoch": 0.7597743629644038,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004947758865194156,
      "loss": 5.2256,
      "mean_token_accuracy": 0.17852199971675872,
      "num_tokens": 16951044.0,
      "step": 9765
    },
    {
      "entropy": 5.646955442428589,
      "epoch": 0.7601633923361214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004947699471487766,
      "loss": 5.3709,
      "mean_token_accuracy": 0.1720644176006317,
      "num_tokens": 16960215.0,
      "step": 9770
    },
    {
      "entropy": 5.659188556671142,
      "epoch": 0.760552421707839,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004947640044434782,
      "loss": 5.4275,
      "mean_token_accuracy": 0.17238795906305313,
      "num_tokens": 16968515.0,
      "step": 9775
    },
    {
      "entropy": 5.6146856307983395,
      "epoch": 0.7609414510795565,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004947580584036109,
      "loss": 5.4029,
      "mean_token_accuracy": 0.16694126427173614,
      "num_tokens": 16977597.0,
      "step": 9780
    },
    {
      "entropy": 5.572262859344482,
      "epoch": 0.761330480451274,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004947521090292649,
      "loss": 5.2411,
      "mean_token_accuracy": 0.17830000072717667,
      "num_tokens": 16987376.0,
      "step": 9785
    },
    {
      "entropy": 5.551488590240479,
      "epoch": 0.7617195098229916,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004947461563205304,
      "loss": 5.2443,
      "mean_token_accuracy": 0.17933500111103057,
      "num_tokens": 16995976.0,
      "step": 9790
    },
    {
      "entropy": 5.513866138458252,
      "epoch": 0.7621085391947092,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004947402002774977,
      "loss": 5.2625,
      "mean_token_accuracy": 0.18817653954029084,
      "num_tokens": 17003819.0,
      "step": 9795
    },
    {
      "entropy": 5.491230010986328,
      "epoch": 0.7624975685664268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004947342409002572,
      "loss": 5.2537,
      "mean_token_accuracy": 0.1792891263961792,
      "num_tokens": 17011632.0,
      "step": 9800
    },
    {
      "entropy": 5.5333374500274655,
      "epoch": 0.7628865979381443,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004947282781888994,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1881726309657097,
      "num_tokens": 17019737.0,
      "step": 9805
    },
    {
      "entropy": 5.583590269088745,
      "epoch": 0.7632756273098619,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004947223121435147,
      "loss": 5.3234,
      "mean_token_accuracy": 0.18121691346168517,
      "num_tokens": 17028803.0,
      "step": 9810
    },
    {
      "entropy": 5.554047870635986,
      "epoch": 0.7636646566815795,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947163427641936,
      "loss": 5.1875,
      "mean_token_accuracy": 0.18053693920373917,
      "num_tokens": 17036847.0,
      "step": 9815
    },
    {
      "entropy": 5.621330976486206,
      "epoch": 0.764053686053297,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004947103700510268,
      "loss": 5.3166,
      "mean_token_accuracy": 0.1728194311261177,
      "num_tokens": 17045665.0,
      "step": 9820
    },
    {
      "entropy": 5.635409545898438,
      "epoch": 0.7644427154250146,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004947043940041047,
      "loss": 5.3207,
      "mean_token_accuracy": 0.18165969252586364,
      "num_tokens": 17053842.0,
      "step": 9825
    },
    {
      "entropy": 5.594769430160523,
      "epoch": 0.7648317447967321,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946984146235183,
      "loss": 5.3953,
      "mean_token_accuracy": 0.16821860522031784,
      "num_tokens": 17062287.0,
      "step": 9830
    },
    {
      "entropy": 5.6085522174835205,
      "epoch": 0.7652207741684497,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946924319093579,
      "loss": 5.383,
      "mean_token_accuracy": 0.17949736267328262,
      "num_tokens": 17071495.0,
      "step": 9835
    },
    {
      "entropy": 5.597342872619629,
      "epoch": 0.7656098035401673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946864458617148,
      "loss": 5.3294,
      "mean_token_accuracy": 0.1748246893286705,
      "num_tokens": 17080734.0,
      "step": 9840
    },
    {
      "entropy": 5.621782112121582,
      "epoch": 0.7659988329118849,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946804564806793,
      "loss": 5.3367,
      "mean_token_accuracy": 0.17630306482315064,
      "num_tokens": 17089683.0,
      "step": 9845
    },
    {
      "entropy": 5.594048404693604,
      "epoch": 0.7663878622836025,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946744637663427,
      "loss": 5.377,
      "mean_token_accuracy": 0.17326258718967438,
      "num_tokens": 17098203.0,
      "step": 9850
    },
    {
      "entropy": 5.5439417362213135,
      "epoch": 0.7667768916553199,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004946684677187956,
      "loss": 5.222,
      "mean_token_accuracy": 0.17714088261127472,
      "num_tokens": 17107185.0,
      "step": 9855
    },
    {
      "entropy": 5.580592250823974,
      "epoch": 0.7671659210270375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946624683381292,
      "loss": 5.2691,
      "mean_token_accuracy": 0.18122341930866243,
      "num_tokens": 17115168.0,
      "step": 9860
    },
    {
      "entropy": 5.545770931243896,
      "epoch": 0.7675549503987551,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004946564656244345,
      "loss": 5.2392,
      "mean_token_accuracy": 0.18823157846927643,
      "num_tokens": 17124037.0,
      "step": 9865
    },
    {
      "entropy": 5.5434150218963625,
      "epoch": 0.7679439797704727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946504595778026,
      "loss": 5.2068,
      "mean_token_accuracy": 0.18442562073469163,
      "num_tokens": 17132238.0,
      "step": 9870
    },
    {
      "entropy": 5.5897815227508545,
      "epoch": 0.7683330091421903,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004946444501983246,
      "loss": 5.3183,
      "mean_token_accuracy": 0.17456935942173005,
      "num_tokens": 17140405.0,
      "step": 9875
    },
    {
      "entropy": 5.550602149963379,
      "epoch": 0.7687220385139077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004946384374860916,
      "loss": 5.2457,
      "mean_token_accuracy": 0.18028399348258972,
      "num_tokens": 17149207.0,
      "step": 9880
    },
    {
      "entropy": 5.561654663085937,
      "epoch": 0.7691110678856253,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004946324214411949,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1710297718644142,
      "num_tokens": 17157680.0,
      "step": 9885
    },
    {
      "entropy": 5.574733448028565,
      "epoch": 0.7695000972573429,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946264020637259,
      "loss": 5.2662,
      "mean_token_accuracy": 0.1798670008778572,
      "num_tokens": 17166907.0,
      "step": 9890
    },
    {
      "entropy": 5.542627191543579,
      "epoch": 0.7698891266290605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946203793537757,
      "loss": 5.3096,
      "mean_token_accuracy": 0.18232055008411407,
      "num_tokens": 17176318.0,
      "step": 9895
    },
    {
      "entropy": 5.583697700500489,
      "epoch": 0.7702781560007781,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946143533114358,
      "loss": 5.3479,
      "mean_token_accuracy": 0.17135487496852875,
      "num_tokens": 17185380.0,
      "step": 9900
    },
    {
      "entropy": 5.611617565155029,
      "epoch": 0.7706671853724957,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004946083239367976,
      "loss": 5.3359,
      "mean_token_accuracy": 0.1717481330037117,
      "num_tokens": 17193177.0,
      "step": 9905
    },
    {
      "entropy": 5.541773319244385,
      "epoch": 0.7710562147442132,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946022912299527,
      "loss": 5.3038,
      "mean_token_accuracy": 0.1723075658082962,
      "num_tokens": 17201071.0,
      "step": 9910
    },
    {
      "entropy": 5.5032923221588135,
      "epoch": 0.7714452441159307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945962551909926,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1849968373775482,
      "num_tokens": 17209375.0,
      "step": 9915
    },
    {
      "entropy": 5.563618803024292,
      "epoch": 0.7718342734876483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945902158200089,
      "loss": 5.1621,
      "mean_token_accuracy": 0.1880572259426117,
      "num_tokens": 17217453.0,
      "step": 9920
    },
    {
      "entropy": 5.600003814697265,
      "epoch": 0.7722233028593659,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945841731170931,
      "loss": 5.2354,
      "mean_token_accuracy": 0.17970992028713226,
      "num_tokens": 17225671.0,
      "step": 9925
    },
    {
      "entropy": 5.5217520236969,
      "epoch": 0.7726123322310835,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004945781270823369,
      "loss": 5.2036,
      "mean_token_accuracy": 0.18175071477890015,
      "num_tokens": 17234396.0,
      "step": 9930
    },
    {
      "entropy": 5.555613374710083,
      "epoch": 0.773001361602801,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004945720777158323,
      "loss": 5.3359,
      "mean_token_accuracy": 0.17705983817577362,
      "num_tokens": 17243509.0,
      "step": 9935
    },
    {
      "entropy": 5.579312658309936,
      "epoch": 0.7733903909745186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945660250176708,
      "loss": 5.2496,
      "mean_token_accuracy": 0.18622122406959535,
      "num_tokens": 17251447.0,
      "step": 9940
    },
    {
      "entropy": 5.535455179214478,
      "epoch": 0.7737794203462361,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004945599689879443,
      "loss": 5.3073,
      "mean_token_accuracy": 0.1836530551314354,
      "num_tokens": 17260240.0,
      "step": 9945
    },
    {
      "entropy": 5.5572467803955075,
      "epoch": 0.7741684497179537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004945539096267448,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1767766907811165,
      "num_tokens": 17267835.0,
      "step": 9950
    },
    {
      "entropy": 5.566209411621093,
      "epoch": 0.7745574790896713,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004945478469341642,
      "loss": 5.3083,
      "mean_token_accuracy": 0.17683991938829421,
      "num_tokens": 17277069.0,
      "step": 9955
    },
    {
      "entropy": 5.575392580032348,
      "epoch": 0.7749465084613888,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945417809102944,
      "loss": 5.293,
      "mean_token_accuracy": 0.1758789226412773,
      "num_tokens": 17285014.0,
      "step": 9960
    },
    {
      "entropy": 5.5071587562561035,
      "epoch": 0.7753355378331064,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004945357115552275,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18336853235960007,
      "num_tokens": 17292649.0,
      "step": 9965
    },
    {
      "entropy": 5.604107475280761,
      "epoch": 0.775724567204824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004945296388690557,
      "loss": 5.3879,
      "mean_token_accuracy": 0.1733226954936981,
      "num_tokens": 17301792.0,
      "step": 9970
    },
    {
      "entropy": 5.540673494338989,
      "epoch": 0.7761135965765416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000494523562851871,
      "loss": 5.1412,
      "mean_token_accuracy": 0.18009239435195923,
      "num_tokens": 17310377.0,
      "step": 9975
    },
    {
      "entropy": 5.568353652954102,
      "epoch": 0.7765026259482591,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004945174835037655,
      "loss": 5.3494,
      "mean_token_accuracy": 0.17051731050014496,
      "num_tokens": 17319779.0,
      "step": 9980
    },
    {
      "entropy": 5.67357063293457,
      "epoch": 0.7768916553199766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004945114008248319,
      "loss": 5.4623,
      "mean_token_accuracy": 0.1699962943792343,
      "num_tokens": 17328210.0,
      "step": 9985
    },
    {
      "entropy": 5.558511638641358,
      "epoch": 0.7772806846916942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945053148151619,
      "loss": 5.2192,
      "mean_token_accuracy": 0.18325317949056624,
      "num_tokens": 17336287.0,
      "step": 9990
    },
    {
      "entropy": 5.536718559265137,
      "epoch": 0.7776697140634118,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944992254748482,
      "loss": 5.2726,
      "mean_token_accuracy": 0.170750629901886,
      "num_tokens": 17344171.0,
      "step": 9995
    },
    {
      "entropy": 5.516432762145996,
      "epoch": 0.7780587434351294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944931328039832,
      "loss": 5.2807,
      "mean_token_accuracy": 0.1826992779970169,
      "num_tokens": 17352786.0,
      "step": 10000
    },
    {
      "entropy": 5.5678215503692625,
      "epoch": 0.778447772806847,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004944870368026592,
      "loss": 5.2731,
      "mean_token_accuracy": 0.1825814127922058,
      "num_tokens": 17361824.0,
      "step": 10005
    },
    {
      "entropy": 5.621897792816162,
      "epoch": 0.7788368021785644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004944809374709689,
      "loss": 5.2789,
      "mean_token_accuracy": 0.18325003683567048,
      "num_tokens": 17370249.0,
      "step": 10010
    },
    {
      "entropy": 5.564912033081055,
      "epoch": 0.779225831550282,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944748348090045,
      "loss": 5.2644,
      "mean_token_accuracy": 0.17794045060873032,
      "num_tokens": 17379650.0,
      "step": 10015
    },
    {
      "entropy": 5.543955659866333,
      "epoch": 0.7796148609219996,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004944687288168589,
      "loss": 5.2846,
      "mean_token_accuracy": 0.17984797060489655,
      "num_tokens": 17388046.0,
      "step": 10020
    },
    {
      "entropy": 5.570498132705689,
      "epoch": 0.7800038902937172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004944626194946246,
      "loss": 5.2616,
      "mean_token_accuracy": 0.1875917673110962,
      "num_tokens": 17395851.0,
      "step": 10025
    },
    {
      "entropy": 5.578702831268311,
      "epoch": 0.7803929196654348,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944565068423945,
      "loss": 5.3197,
      "mean_token_accuracy": 0.17798285335302352,
      "num_tokens": 17404273.0,
      "step": 10030
    },
    {
      "entropy": 5.5402006149292,
      "epoch": 0.7807819490371523,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944503908602612,
      "loss": 5.2692,
      "mean_token_accuracy": 0.180065880715847,
      "num_tokens": 17412884.0,
      "step": 10035
    },
    {
      "entropy": 5.602862596511841,
      "epoch": 0.7811709784088698,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944442715483174,
      "loss": 5.2594,
      "mean_token_accuracy": 0.17478689551353455,
      "num_tokens": 17420939.0,
      "step": 10040
    },
    {
      "entropy": 5.561760759353637,
      "epoch": 0.7815600077805874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004944381489066561,
      "loss": 5.3074,
      "mean_token_accuracy": 0.1774538889527321,
      "num_tokens": 17429838.0,
      "step": 10045
    },
    {
      "entropy": 5.557232141494751,
      "epoch": 0.781949037152305,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004944320229353702,
      "loss": 5.2628,
      "mean_token_accuracy": 0.17559466063976287,
      "num_tokens": 17439164.0,
      "step": 10050
    },
    {
      "entropy": 5.619220304489136,
      "epoch": 0.7823380665240226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004944258936345526,
      "loss": 5.2689,
      "mean_token_accuracy": 0.18301373720169067,
      "num_tokens": 17447544.0,
      "step": 10055
    },
    {
      "entropy": 5.492280721664429,
      "epoch": 0.7827270958957401,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944197610042963,
      "loss": 5.1387,
      "mean_token_accuracy": 0.1840054675936699,
      "num_tokens": 17456270.0,
      "step": 10060
    },
    {
      "entropy": 5.554314136505127,
      "epoch": 0.7831161252674577,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944136250446944,
      "loss": 5.3176,
      "mean_token_accuracy": 0.17473938763141633,
      "num_tokens": 17465241.0,
      "step": 10065
    },
    {
      "entropy": 5.646536111831665,
      "epoch": 0.7835051546391752,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944074857558399,
      "loss": 5.3438,
      "mean_token_accuracy": 0.1689807027578354,
      "num_tokens": 17474460.0,
      "step": 10070
    },
    {
      "entropy": 5.56577320098877,
      "epoch": 0.7838941840108928,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944013431378261,
      "loss": 5.2937,
      "mean_token_accuracy": 0.18041094094514848,
      "num_tokens": 17483256.0,
      "step": 10075
    },
    {
      "entropy": 5.6156011581420895,
      "epoch": 0.7842832133826104,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004943951971907461,
      "loss": 5.3056,
      "mean_token_accuracy": 0.17655237317085265,
      "num_tokens": 17491922.0,
      "step": 10080
    },
    {
      "entropy": 5.613048505783081,
      "epoch": 0.7846722427543279,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943890479146933,
      "loss": 5.2938,
      "mean_token_accuracy": 0.18363881558179856,
      "num_tokens": 17499763.0,
      "step": 10085
    },
    {
      "entropy": 5.548869562149048,
      "epoch": 0.7850612721260455,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943828953097607,
      "loss": 5.2761,
      "mean_token_accuracy": 0.17813890874385835,
      "num_tokens": 17508682.0,
      "step": 10090
    },
    {
      "entropy": 5.520614910125732,
      "epoch": 0.7854503014977631,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494376739376042,
      "loss": 5.2158,
      "mean_token_accuracy": 0.1830417662858963,
      "num_tokens": 17517269.0,
      "step": 10095
    },
    {
      "entropy": 5.505959224700928,
      "epoch": 0.7858393308694807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943705801136304,
      "loss": 5.2342,
      "mean_token_accuracy": 0.17939852327108383,
      "num_tokens": 17525135.0,
      "step": 10100
    },
    {
      "entropy": 5.555085182189941,
      "epoch": 0.7862283602411982,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004943644175226193,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18469279259443283,
      "num_tokens": 17534195.0,
      "step": 10105
    },
    {
      "entropy": 5.541886472702027,
      "epoch": 0.7866173896129158,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004943582516031025,
      "loss": 5.2105,
      "mean_token_accuracy": 0.18273035436868668,
      "num_tokens": 17543016.0,
      "step": 10110
    },
    {
      "entropy": 5.5975395202636715,
      "epoch": 0.7870064189846333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943520823551732,
      "loss": 5.3202,
      "mean_token_accuracy": 0.18122852444648743,
      "num_tokens": 17551854.0,
      "step": 10115
    },
    {
      "entropy": 5.5274608612060545,
      "epoch": 0.7873954483563509,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004943459097789252,
      "loss": 5.2235,
      "mean_token_accuracy": 0.1863451659679413,
      "num_tokens": 17560739.0,
      "step": 10120
    },
    {
      "entropy": 5.535266923904419,
      "epoch": 0.7877844777280685,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943397338744523,
      "loss": 5.2287,
      "mean_token_accuracy": 0.18101687282323836,
      "num_tokens": 17569784.0,
      "step": 10125
    },
    {
      "entropy": 5.592965888977051,
      "epoch": 0.7881735070997861,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494333554641848,
      "loss": 5.3998,
      "mean_token_accuracy": 0.1753152623772621,
      "num_tokens": 17579262.0,
      "step": 10130
    },
    {
      "entropy": 5.508724594116211,
      "epoch": 0.7885625364715037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494327372081206,
      "loss": 5.2582,
      "mean_token_accuracy": 0.1788112297654152,
      "num_tokens": 17587984.0,
      "step": 10135
    },
    {
      "entropy": 5.50737943649292,
      "epoch": 0.7889515658432211,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943211861926204,
      "loss": 5.2049,
      "mean_token_accuracy": 0.18378990292549133,
      "num_tokens": 17596598.0,
      "step": 10140
    },
    {
      "entropy": 5.536486291885376,
      "epoch": 0.7893405952149387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943149969761848,
      "loss": 5.3386,
      "mean_token_accuracy": 0.1783975303173065,
      "num_tokens": 17605768.0,
      "step": 10145
    },
    {
      "entropy": 5.597452116012573,
      "epoch": 0.7897296245866563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943088044319932,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1779785394668579,
      "num_tokens": 17614412.0,
      "step": 10150
    },
    {
      "entropy": 5.5843816757202145,
      "epoch": 0.7901186539583739,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004943026085601395,
      "loss": 5.2841,
      "mean_token_accuracy": 0.18136109709739684,
      "num_tokens": 17622515.0,
      "step": 10155
    },
    {
      "entropy": 5.504694604873658,
      "epoch": 0.7905076833300915,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942964093607178,
      "loss": 5.2068,
      "mean_token_accuracy": 0.18027411699295043,
      "num_tokens": 17631256.0,
      "step": 10160
    },
    {
      "entropy": 5.62631516456604,
      "epoch": 0.790896712701809,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004942902068338222,
      "loss": 5.3251,
      "mean_token_accuracy": 0.17326385974884034,
      "num_tokens": 17639648.0,
      "step": 10165
    },
    {
      "entropy": 5.5855039119720455,
      "epoch": 0.7912857420735265,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942840009795466,
      "loss": 5.2646,
      "mean_token_accuracy": 0.17888859510421753,
      "num_tokens": 17648225.0,
      "step": 10170
    },
    {
      "entropy": 5.58572039604187,
      "epoch": 0.7916747714452441,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942777917979855,
      "loss": 5.2949,
      "mean_token_accuracy": 0.1864625632762909,
      "num_tokens": 17656764.0,
      "step": 10175
    },
    {
      "entropy": 5.564801359176636,
      "epoch": 0.7920638008169617,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004942715792892328,
      "loss": 5.282,
      "mean_token_accuracy": 0.17569558769464494,
      "num_tokens": 17665127.0,
      "step": 10180
    },
    {
      "entropy": 5.501125431060791,
      "epoch": 0.7924528301886793,
      "grad_norm": 1.078125,
      "learning_rate": 0.000494265363453383,
      "loss": 5.167,
      "mean_token_accuracy": 0.1871642664074898,
      "num_tokens": 17673297.0,
      "step": 10185
    },
    {
      "entropy": 5.53309645652771,
      "epoch": 0.7928418595603968,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942591442905302,
      "loss": 5.3145,
      "mean_token_accuracy": 0.1781333178281784,
      "num_tokens": 17681816.0,
      "step": 10190
    },
    {
      "entropy": 5.578197336196899,
      "epoch": 0.7932308889321144,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004942529218007689,
      "loss": 5.3235,
      "mean_token_accuracy": 0.17432713359594346,
      "num_tokens": 17690588.0,
      "step": 10195
    },
    {
      "entropy": 5.577566242218017,
      "epoch": 0.7936199183038319,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942466959841936,
      "loss": 5.2294,
      "mean_token_accuracy": 0.18250014781951904,
      "num_tokens": 17699196.0,
      "step": 10200
    },
    {
      "entropy": 5.604201364517212,
      "epoch": 0.7940089476755495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942404668408985,
      "loss": 5.3432,
      "mean_token_accuracy": 0.17931440323591233,
      "num_tokens": 17708928.0,
      "step": 10205
    },
    {
      "entropy": 5.599688148498535,
      "epoch": 0.7943979770472671,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004942342343709783,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1876217842102051,
      "num_tokens": 17717888.0,
      "step": 10210
    },
    {
      "entropy": 5.620655679702759,
      "epoch": 0.7947870064189846,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942279985745275,
      "loss": 5.2632,
      "mean_token_accuracy": 0.17752481251955032,
      "num_tokens": 17726022.0,
      "step": 10215
    },
    {
      "entropy": 5.512254524230957,
      "epoch": 0.7951760357907022,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004942217594516409,
      "loss": 5.2353,
      "mean_token_accuracy": 0.18148309588432313,
      "num_tokens": 17733998.0,
      "step": 10220
    },
    {
      "entropy": 5.602106142044067,
      "epoch": 0.7955650651624198,
      "grad_norm": 1.09375,
      "learning_rate": 0.000494215517002413,
      "loss": 5.2989,
      "mean_token_accuracy": 0.1809411719441414,
      "num_tokens": 17742039.0,
      "step": 10225
    },
    {
      "entropy": 5.5189722061157225,
      "epoch": 0.7959540945341373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942092712269384,
      "loss": 5.1802,
      "mean_token_accuracy": 0.18697760850191117,
      "num_tokens": 17751474.0,
      "step": 10230
    },
    {
      "entropy": 5.505219650268555,
      "epoch": 0.7963431239058549,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004942030221253122,
      "loss": 5.2039,
      "mean_token_accuracy": 0.18463034331798553,
      "num_tokens": 17759483.0,
      "step": 10235
    },
    {
      "entropy": 5.534953975677491,
      "epoch": 0.7967321532775724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941967696976289,
      "loss": 5.3131,
      "mean_token_accuracy": 0.18131082355976105,
      "num_tokens": 17768025.0,
      "step": 10240
    },
    {
      "entropy": 5.512156438827515,
      "epoch": 0.79712118264929,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004941905139439835,
      "loss": 5.1613,
      "mean_token_accuracy": 0.1847629204392433,
      "num_tokens": 17776861.0,
      "step": 10245
    },
    {
      "entropy": 5.5302856922149655,
      "epoch": 0.7975102120210076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004941842548644712,
      "loss": 5.2737,
      "mean_token_accuracy": 0.1749660238623619,
      "num_tokens": 17786221.0,
      "step": 10250
    },
    {
      "entropy": 5.623866415023803,
      "epoch": 0.7978992413927252,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004941779924591864,
      "loss": 5.2684,
      "mean_token_accuracy": 0.17311887294054032,
      "num_tokens": 17794619.0,
      "step": 10255
    },
    {
      "entropy": 5.59625391960144,
      "epoch": 0.7982882707644428,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941717267282244,
      "loss": 5.261,
      "mean_token_accuracy": 0.1823060542345047,
      "num_tokens": 17803288.0,
      "step": 10260
    },
    {
      "entropy": 5.552209997177124,
      "epoch": 0.7986773001361602,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941654576716806,
      "loss": 5.3259,
      "mean_token_accuracy": 0.16714696437120438,
      "num_tokens": 17812648.0,
      "step": 10265
    },
    {
      "entropy": 5.487466621398926,
      "epoch": 0.7990663295078778,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941591852896495,
      "loss": 5.2442,
      "mean_token_accuracy": 0.1840747281908989,
      "num_tokens": 17821955.0,
      "step": 10270
    },
    {
      "entropy": 5.648491287231446,
      "epoch": 0.7994553588795954,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004941529095822268,
      "loss": 5.3107,
      "mean_token_accuracy": 0.18278683423995973,
      "num_tokens": 17829987.0,
      "step": 10275
    },
    {
      "entropy": 5.578384351730347,
      "epoch": 0.799844388251313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941466305495074,
      "loss": 5.2332,
      "mean_token_accuracy": 0.18437684029340745,
      "num_tokens": 17839298.0,
      "step": 10280
    },
    {
      "entropy": 5.54558835029602,
      "epoch": 0.8002334176230306,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941403481915867,
      "loss": 5.2554,
      "mean_token_accuracy": 0.17760299891233444,
      "num_tokens": 17847950.0,
      "step": 10285
    },
    {
      "entropy": 5.613780212402344,
      "epoch": 0.800622446994748,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004941340625085601,
      "loss": 5.3839,
      "mean_token_accuracy": 0.17219481319189073,
      "num_tokens": 17857416.0,
      "step": 10290
    },
    {
      "entropy": 5.634960556030274,
      "epoch": 0.8010114763664656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004941277735005228,
      "loss": 5.2557,
      "mean_token_accuracy": 0.18916381150484085,
      "num_tokens": 17866516.0,
      "step": 10295
    },
    {
      "entropy": 5.681487607955932,
      "epoch": 0.8014005057381832,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941214811675702,
      "loss": 5.4014,
      "mean_token_accuracy": 0.16773732006549835,
      "num_tokens": 17875823.0,
      "step": 10300
    },
    {
      "entropy": 5.512523746490478,
      "epoch": 0.8017895351099008,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004941151855097982,
      "loss": 5.2769,
      "mean_token_accuracy": 0.17694343477487565,
      "num_tokens": 17883597.0,
      "step": 10305
    },
    {
      "entropy": 5.593393468856812,
      "epoch": 0.8021785644816184,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004941088865273018,
      "loss": 5.2708,
      "mean_token_accuracy": 0.17596012204885483,
      "num_tokens": 17892053.0,
      "step": 10310
    },
    {
      "entropy": 5.583527565002441,
      "epoch": 0.802567593853336,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941025842201769,
      "loss": 5.1833,
      "mean_token_accuracy": 0.18942164480686188,
      "num_tokens": 17900187.0,
      "step": 10315
    },
    {
      "entropy": 5.563230037689209,
      "epoch": 0.8029566232250535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004940962785885189,
      "loss": 5.3392,
      "mean_token_accuracy": 0.1744835615158081,
      "num_tokens": 17908608.0,
      "step": 10320
    },
    {
      "entropy": 5.588981246948242,
      "epoch": 0.803345652596771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940899696324237,
      "loss": 5.2302,
      "mean_token_accuracy": 0.18067265748977662,
      "num_tokens": 17917331.0,
      "step": 10325
    },
    {
      "entropy": 5.626565408706665,
      "epoch": 0.8037346819684886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004940836573519868,
      "loss": 5.3506,
      "mean_token_accuracy": 0.17576338946819306,
      "num_tokens": 17925228.0,
      "step": 10330
    },
    {
      "entropy": 5.5760797500610355,
      "epoch": 0.8041237113402062,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004940773417473043,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1828831896185875,
      "num_tokens": 17934110.0,
      "step": 10335
    },
    {
      "entropy": 5.489350032806397,
      "epoch": 0.8045127407119238,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004940710228184717,
      "loss": 5.2238,
      "mean_token_accuracy": 0.18484015315771102,
      "num_tokens": 17942237.0,
      "step": 10340
    },
    {
      "entropy": 5.491905260086059,
      "epoch": 0.8049017700836413,
      "grad_norm": 1.046875,
      "learning_rate": 0.000494064700565585,
      "loss": 5.0699,
      "mean_token_accuracy": 0.1908535286784172,
      "num_tokens": 17951288.0,
      "step": 10345
    },
    {
      "entropy": 5.618882369995117,
      "epoch": 0.8052907994553589,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004940583749887403,
      "loss": 5.3416,
      "mean_token_accuracy": 0.17543888986110687,
      "num_tokens": 17960300.0,
      "step": 10350
    },
    {
      "entropy": 5.532686710357666,
      "epoch": 0.8056798288270764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940520460880333,
      "loss": 5.2459,
      "mean_token_accuracy": 0.18015871644020082,
      "num_tokens": 17968745.0,
      "step": 10355
    },
    {
      "entropy": 5.501589345932007,
      "epoch": 0.806068858198794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940457138635601,
      "loss": 5.2277,
      "mean_token_accuracy": 0.18703533113002777,
      "num_tokens": 17976614.0,
      "step": 10360
    },
    {
      "entropy": 5.590014410018921,
      "epoch": 0.8064578875705116,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004940393783154169,
      "loss": 5.3707,
      "mean_token_accuracy": 0.17463206201791764,
      "num_tokens": 17984769.0,
      "step": 10365
    },
    {
      "entropy": 5.55709285736084,
      "epoch": 0.8068469169422291,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004940330394436999,
      "loss": 5.2467,
      "mean_token_accuracy": 0.1752379283308983,
      "num_tokens": 17992885.0,
      "step": 10370
    },
    {
      "entropy": 5.605456304550171,
      "epoch": 0.8072359463139467,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494026697248505,
      "loss": 5.2807,
      "mean_token_accuracy": 0.17333069145679475,
      "num_tokens": 18001752.0,
      "step": 10375
    },
    {
      "entropy": 5.528792381286621,
      "epoch": 0.8076249756856643,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004940203517299287,
      "loss": 5.2895,
      "mean_token_accuracy": 0.18906021267175674,
      "num_tokens": 18009939.0,
      "step": 10380
    },
    {
      "entropy": 5.631753206253052,
      "epoch": 0.8080140050573819,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000494014002888067,
      "loss": 5.2464,
      "mean_token_accuracy": 0.1893475979566574,
      "num_tokens": 18017993.0,
      "step": 10385
    },
    {
      "entropy": 5.656357955932617,
      "epoch": 0.8084030344290994,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940076507230166,
      "loss": 5.3562,
      "mean_token_accuracy": 0.1703679695725441,
      "num_tokens": 18027564.0,
      "step": 10390
    },
    {
      "entropy": 5.6072712421417235,
      "epoch": 0.8087920638008169,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004940012952348735,
      "loss": 5.3296,
      "mean_token_accuracy": 0.17833323925733566,
      "num_tokens": 18036174.0,
      "step": 10395
    },
    {
      "entropy": 5.577934980392456,
      "epoch": 0.8091810931725345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004939949364237345,
      "loss": 5.2453,
      "mean_token_accuracy": 0.1775733470916748,
      "num_tokens": 18044994.0,
      "step": 10400
    },
    {
      "entropy": 5.5368577003479,
      "epoch": 0.8095701225442521,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004939885742896958,
      "loss": 5.2578,
      "mean_token_accuracy": 0.18077761679887772,
      "num_tokens": 18053789.0,
      "step": 10405
    },
    {
      "entropy": 5.596770715713501,
      "epoch": 0.8099591519159697,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000493982208832854,
      "loss": 5.3559,
      "mean_token_accuracy": 0.1795223593711853,
      "num_tokens": 18063483.0,
      "step": 10410
    },
    {
      "entropy": 5.554763412475586,
      "epoch": 0.8103481812876873,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939758400533058,
      "loss": 5.1755,
      "mean_token_accuracy": 0.1859259694814682,
      "num_tokens": 18071952.0,
      "step": 10415
    },
    {
      "entropy": 5.574611568450928,
      "epoch": 0.8107372106594047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004939694679511478,
      "loss": 5.2283,
      "mean_token_accuracy": 0.18664990812540055,
      "num_tokens": 18080323.0,
      "step": 10420
    },
    {
      "entropy": 5.606487798690796,
      "epoch": 0.8111262400311223,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939630925264765,
      "loss": 5.3384,
      "mean_token_accuracy": 0.17754612863063812,
      "num_tokens": 18089458.0,
      "step": 10425
    },
    {
      "entropy": 5.600635194778443,
      "epoch": 0.8115152694028399,
      "grad_norm": 1.0,
      "learning_rate": 0.000493956713779389,
      "loss": 5.2474,
      "mean_token_accuracy": 0.1756208136677742,
      "num_tokens": 18098553.0,
      "step": 10430
    },
    {
      "entropy": 5.55400333404541,
      "epoch": 0.8119042987745575,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939503317099817,
      "loss": 5.357,
      "mean_token_accuracy": 0.17223884910345078,
      "num_tokens": 18106794.0,
      "step": 10435
    },
    {
      "entropy": 5.533067417144776,
      "epoch": 0.8122933281462751,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004939439463183517,
      "loss": 5.1949,
      "mean_token_accuracy": 0.1866455078125,
      "num_tokens": 18115509.0,
      "step": 10440
    },
    {
      "entropy": 5.5484973907470705,
      "epoch": 0.8126823575179926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939375576045958,
      "loss": 5.3523,
      "mean_token_accuracy": 0.1700367271900177,
      "num_tokens": 18124548.0,
      "step": 10445
    },
    {
      "entropy": 5.563609457015991,
      "epoch": 0.8130713868897101,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004939311655688109,
      "loss": 5.2722,
      "mean_token_accuracy": 0.17956678569316864,
      "num_tokens": 18132295.0,
      "step": 10450
    },
    {
      "entropy": 5.647200250625611,
      "epoch": 0.8134604162614277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004939247702110941,
      "loss": 5.3737,
      "mean_token_accuracy": 0.17452162355184556,
      "num_tokens": 18140731.0,
      "step": 10455
    },
    {
      "entropy": 5.566962099075317,
      "epoch": 0.8138494456331453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939183715315423,
      "loss": 5.2356,
      "mean_token_accuracy": 0.17991308867931366,
      "num_tokens": 18149834.0,
      "step": 10460
    },
    {
      "entropy": 5.526085138320923,
      "epoch": 0.8142384750048629,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004939119695302526,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19115205854177475,
      "num_tokens": 18158745.0,
      "step": 10465
    },
    {
      "entropy": 5.607345056533814,
      "epoch": 0.8146275043765804,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939055642073224,
      "loss": 5.3817,
      "mean_token_accuracy": 0.17092738747596742,
      "num_tokens": 18168292.0,
      "step": 10470
    },
    {
      "entropy": 5.592886638641358,
      "epoch": 0.815016533748298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004938991555628484,
      "loss": 5.2716,
      "mean_token_accuracy": 0.17934371829032897,
      "num_tokens": 18177592.0,
      "step": 10475
    },
    {
      "entropy": 5.5881054401397705,
      "epoch": 0.8154055631200156,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004938927435969283,
      "loss": 5.2454,
      "mean_token_accuracy": 0.17746430188417434,
      "num_tokens": 18186845.0,
      "step": 10480
    },
    {
      "entropy": 5.613975381851196,
      "epoch": 0.8157945924917331,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938863283096592,
      "loss": 5.3401,
      "mean_token_accuracy": 0.17624317705631257,
      "num_tokens": 18195417.0,
      "step": 10485
    },
    {
      "entropy": 5.5749125480651855,
      "epoch": 0.8161836218634507,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004938799097011384,
      "loss": 5.2681,
      "mean_token_accuracy": 0.17838387191295624,
      "num_tokens": 18204009.0,
      "step": 10490
    },
    {
      "entropy": 5.531589126586914,
      "epoch": 0.8165726512351682,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004938734877714634,
      "loss": 5.2806,
      "mean_token_accuracy": 0.18555265069007873,
      "num_tokens": 18211927.0,
      "step": 10495
    },
    {
      "entropy": 5.482576751708985,
      "epoch": 0.8169616806068858,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004938670625207317,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1902029275894165,
      "num_tokens": 18220394.0,
      "step": 10500
    },
    {
      "entropy": 5.596595764160156,
      "epoch": 0.8173507099786034,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004938606339490406,
      "loss": 5.3268,
      "mean_token_accuracy": 0.17240248173475264,
      "num_tokens": 18229803.0,
      "step": 10505
    },
    {
      "entropy": 5.591364812850952,
      "epoch": 0.817739739350321,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004938542020564877,
      "loss": 5.25,
      "mean_token_accuracy": 0.1745932385325432,
      "num_tokens": 18238431.0,
      "step": 10510
    },
    {
      "entropy": 5.563290119171143,
      "epoch": 0.8181287687220385,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004938477668431706,
      "loss": 5.2631,
      "mean_token_accuracy": 0.1808483213186264,
      "num_tokens": 18246745.0,
      "step": 10515
    },
    {
      "entropy": 5.531867933273316,
      "epoch": 0.8185177980937561,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004938413283091871,
      "loss": 5.2347,
      "mean_token_accuracy": 0.1842041239142418,
      "num_tokens": 18255569.0,
      "step": 10520
    },
    {
      "entropy": 5.529405164718628,
      "epoch": 0.8189068274654736,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004938348864546347,
      "loss": 5.1893,
      "mean_token_accuracy": 0.18322349041700364,
      "num_tokens": 18263711.0,
      "step": 10525
    },
    {
      "entropy": 5.641204309463501,
      "epoch": 0.8192958568371912,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938284412796111,
      "loss": 5.3469,
      "mean_token_accuracy": 0.1820904418826103,
      "num_tokens": 18272220.0,
      "step": 10530
    },
    {
      "entropy": 5.606492519378662,
      "epoch": 0.8196848862089088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004938219927842144,
      "loss": 5.2645,
      "mean_token_accuracy": 0.17838217914104462,
      "num_tokens": 18281107.0,
      "step": 10535
    },
    {
      "entropy": 5.482708406448364,
      "epoch": 0.8200739155806264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938155409685422,
      "loss": 5.217,
      "mean_token_accuracy": 0.18029659390449523,
      "num_tokens": 18290196.0,
      "step": 10540
    },
    {
      "entropy": 5.607309103012085,
      "epoch": 0.820462944952344,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938090858326923,
      "loss": 5.3763,
      "mean_token_accuracy": 0.17410530298948287,
      "num_tokens": 18299418.0,
      "step": 10545
    },
    {
      "entropy": 5.490515804290771,
      "epoch": 0.8208519743240614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493802627376763,
      "loss": 5.1043,
      "mean_token_accuracy": 0.19295817613601685,
      "num_tokens": 18308171.0,
      "step": 10550
    },
    {
      "entropy": 5.5814563751220705,
      "epoch": 0.821241003695779,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937961656008518,
      "loss": 5.2829,
      "mean_token_accuracy": 0.18063021302223206,
      "num_tokens": 18317144.0,
      "step": 10555
    },
    {
      "entropy": 5.5784365177154545,
      "epoch": 0.8216300330674966,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004937897005050573,
      "loss": 5.3426,
      "mean_token_accuracy": 0.17652396112680435,
      "num_tokens": 18326648.0,
      "step": 10560
    },
    {
      "entropy": 5.636283016204834,
      "epoch": 0.8220190624392142,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004937832320894772,
      "loss": 5.312,
      "mean_token_accuracy": 0.18171336203813554,
      "num_tokens": 18335901.0,
      "step": 10565
    },
    {
      "entropy": 5.566133880615235,
      "epoch": 0.8224080918109318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004937767603542098,
      "loss": 5.2277,
      "mean_token_accuracy": 0.18326659202575685,
      "num_tokens": 18344905.0,
      "step": 10570
    },
    {
      "entropy": 5.479534387588501,
      "epoch": 0.8227971211826492,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004937702852993534,
      "loss": 5.1877,
      "mean_token_accuracy": 0.18900362253189087,
      "num_tokens": 18353422.0,
      "step": 10575
    },
    {
      "entropy": 5.695352077484131,
      "epoch": 0.8231861505543668,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000493763806925006,
      "loss": 5.3623,
      "mean_token_accuracy": 0.1664465293288231,
      "num_tokens": 18361632.0,
      "step": 10580
    },
    {
      "entropy": 5.53762264251709,
      "epoch": 0.8235751799260844,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004937573252312661,
      "loss": 5.1894,
      "mean_token_accuracy": 0.19036249071359634,
      "num_tokens": 18370142.0,
      "step": 10585
    },
    {
      "entropy": 5.4955779075622555,
      "epoch": 0.823964209297802,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493750840218232,
      "loss": 5.2349,
      "mean_token_accuracy": 0.18061405271291733,
      "num_tokens": 18378245.0,
      "step": 10590
    },
    {
      "entropy": 5.531733751296997,
      "epoch": 0.8243532386695196,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937443518860021,
      "loss": 5.1946,
      "mean_token_accuracy": 0.1911581426858902,
      "num_tokens": 18386814.0,
      "step": 10595
    },
    {
      "entropy": 5.556150722503662,
      "epoch": 0.8247422680412371,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004937378602346747,
      "loss": 5.2209,
      "mean_token_accuracy": 0.18298658430576326,
      "num_tokens": 18396276.0,
      "step": 10600
    },
    {
      "entropy": 5.552563619613648,
      "epoch": 0.8251312974129547,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004937313652643485,
      "loss": 5.2183,
      "mean_token_accuracy": 0.18201106637716294,
      "num_tokens": 18404445.0,
      "step": 10605
    },
    {
      "entropy": 5.528123998641968,
      "epoch": 0.8255203267846722,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493724866975122,
      "loss": 5.1337,
      "mean_token_accuracy": 0.1917639896273613,
      "num_tokens": 18412870.0,
      "step": 10610
    },
    {
      "entropy": 5.5436375617980955,
      "epoch": 0.8259093561563898,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937183653670937,
      "loss": 5.2595,
      "mean_token_accuracy": 0.18410644680261612,
      "num_tokens": 18421084.0,
      "step": 10615
    },
    {
      "entropy": 5.524842262268066,
      "epoch": 0.8262983855281074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937118604403623,
      "loss": 5.2877,
      "mean_token_accuracy": 0.17960087060928345,
      "num_tokens": 18429869.0,
      "step": 10620
    },
    {
      "entropy": 5.540694093704223,
      "epoch": 0.8266874148998249,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004937053521950266,
      "loss": 5.1499,
      "mean_token_accuracy": 0.1879689648747444,
      "num_tokens": 18438747.0,
      "step": 10625
    },
    {
      "entropy": 5.6041404724121096,
      "epoch": 0.8270764442715425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936988406311854,
      "loss": 5.3532,
      "mean_token_accuracy": 0.17607826888561248,
      "num_tokens": 18447241.0,
      "step": 10630
    },
    {
      "entropy": 5.625930404663086,
      "epoch": 0.8274654736432601,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936923257489373,
      "loss": 5.3744,
      "mean_token_accuracy": 0.17208832800388335,
      "num_tokens": 18455757.0,
      "step": 10635
    },
    {
      "entropy": 5.518155956268311,
      "epoch": 0.8278545030149776,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004936858075483811,
      "loss": 5.211,
      "mean_token_accuracy": 0.1866015300154686,
      "num_tokens": 18463863.0,
      "step": 10640
    },
    {
      "entropy": 5.5533936500549315,
      "epoch": 0.8282435323866952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493679286029616,
      "loss": 5.3891,
      "mean_token_accuracy": 0.17119386345148085,
      "num_tokens": 18472396.0,
      "step": 10645
    },
    {
      "entropy": 5.640747547149658,
      "epoch": 0.8286325617584127,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004936727611927407,
      "loss": 5.2427,
      "mean_token_accuracy": 0.175209204852581,
      "num_tokens": 18480625.0,
      "step": 10650
    },
    {
      "entropy": 5.61053056716919,
      "epoch": 0.8290215911301303,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936662330378546,
      "loss": 5.3087,
      "mean_token_accuracy": 0.1735701560974121,
      "num_tokens": 18489878.0,
      "step": 10655
    },
    {
      "entropy": 5.51786322593689,
      "epoch": 0.8294106205018479,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004936597015650561,
      "loss": 5.2001,
      "mean_token_accuracy": 0.1842064917087555,
      "num_tokens": 18498873.0,
      "step": 10660
    },
    {
      "entropy": 5.62124195098877,
      "epoch": 0.8297996498735655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004936531667744448,
      "loss": 5.3311,
      "mean_token_accuracy": 0.17280071675777436,
      "num_tokens": 18507716.0,
      "step": 10665
    },
    {
      "entropy": 5.5693971633911135,
      "epoch": 0.8301886792452831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936466286661197,
      "loss": 5.1568,
      "mean_token_accuracy": 0.19043404161930083,
      "num_tokens": 18516547.0,
      "step": 10670
    },
    {
      "entropy": 5.5463659286499025,
      "epoch": 0.8305777086170005,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049364008724018,
      "loss": 5.329,
      "mean_token_accuracy": 0.17302707582712173,
      "num_tokens": 18525714.0,
      "step": 10675
    },
    {
      "entropy": 5.606787633895874,
      "epoch": 0.8309667379887181,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493633542496725,
      "loss": 5.3393,
      "mean_token_accuracy": 0.18241952061653138,
      "num_tokens": 18534981.0,
      "step": 10680
    },
    {
      "entropy": 5.614515066146851,
      "epoch": 0.8313557673604357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004936269944358539,
      "loss": 5.2568,
      "mean_token_accuracy": 0.18250710964202882,
      "num_tokens": 18543599.0,
      "step": 10685
    },
    {
      "entropy": 5.5610284328460695,
      "epoch": 0.8317447967321533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004936204430576664,
      "loss": 5.2429,
      "mean_token_accuracy": 0.18756332397460937,
      "num_tokens": 18552207.0,
      "step": 10690
    },
    {
      "entropy": 5.532431745529175,
      "epoch": 0.8321338261038709,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004936138883622614,
      "loss": 5.2997,
      "mean_token_accuracy": 0.17762846350669861,
      "num_tokens": 18560421.0,
      "step": 10695
    },
    {
      "entropy": 5.552207374572754,
      "epoch": 0.8325228554755884,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004936073303497387,
      "loss": 5.1879,
      "mean_token_accuracy": 0.19418384432792662,
      "num_tokens": 18568812.0,
      "step": 10700
    },
    {
      "entropy": 5.597041654586792,
      "epoch": 0.8329118848473059,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936007690201976,
      "loss": 5.1809,
      "mean_token_accuracy": 0.18028334379196168,
      "num_tokens": 18577077.0,
      "step": 10705
    },
    {
      "entropy": 5.587600755691528,
      "epoch": 0.8333009142190235,
      "grad_norm": 1.015625,
      "learning_rate": 0.000493594204373738,
      "loss": 5.3063,
      "mean_token_accuracy": 0.18268045634031296,
      "num_tokens": 18585754.0,
      "step": 10710
    },
    {
      "entropy": 5.5249193668365475,
      "epoch": 0.8336899435907411,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004935876364104591,
      "loss": 5.1758,
      "mean_token_accuracy": 0.18650716692209243,
      "num_tokens": 18594578.0,
      "step": 10715
    },
    {
      "entropy": 5.610929346084594,
      "epoch": 0.8340789729624587,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935810651304608,
      "loss": 5.2808,
      "mean_token_accuracy": 0.17683386504650117,
      "num_tokens": 18603810.0,
      "step": 10720
    },
    {
      "entropy": 5.6428618907928465,
      "epoch": 0.8344680023341763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004935744905338427,
      "loss": 5.2707,
      "mean_token_accuracy": 0.1749935641884804,
      "num_tokens": 18612870.0,
      "step": 10725
    },
    {
      "entropy": 5.568986654281616,
      "epoch": 0.8348570317058938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004935679126207046,
      "loss": 5.321,
      "mean_token_accuracy": 0.183121457695961,
      "num_tokens": 18621353.0,
      "step": 10730
    },
    {
      "entropy": 5.484223747253418,
      "epoch": 0.8352460610776113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935613313911463,
      "loss": 5.1993,
      "mean_token_accuracy": 0.18471705168485641,
      "num_tokens": 18630152.0,
      "step": 10735
    },
    {
      "entropy": 5.574936437606811,
      "epoch": 0.8356350904493289,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004935547468452678,
      "loss": 5.2353,
      "mean_token_accuracy": 0.18468862175941467,
      "num_tokens": 18639278.0,
      "step": 10740
    },
    {
      "entropy": 5.602726650238037,
      "epoch": 0.8360241198210465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004935481589831688,
      "loss": 5.2411,
      "mean_token_accuracy": 0.18399454802274703,
      "num_tokens": 18647436.0,
      "step": 10745
    },
    {
      "entropy": 5.6336143016815186,
      "epoch": 0.8364131491927641,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935415678049492,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1739118590950966,
      "num_tokens": 18655532.0,
      "step": 10750
    },
    {
      "entropy": 5.601388788223266,
      "epoch": 0.8368021785644816,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935349733107094,
      "loss": 5.2894,
      "mean_token_accuracy": 0.1812187984585762,
      "num_tokens": 18664827.0,
      "step": 10755
    },
    {
      "entropy": 5.541911458969116,
      "epoch": 0.8371912079361992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493528375500549,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1848923996090889,
      "num_tokens": 18673535.0,
      "step": 10760
    },
    {
      "entropy": 5.552118492126465,
      "epoch": 0.8375802373079168,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935217743745685,
      "loss": 5.3898,
      "mean_token_accuracy": 0.16902155578136444,
      "num_tokens": 18682223.0,
      "step": 10765
    },
    {
      "entropy": 5.61554102897644,
      "epoch": 0.8379692666796343,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004935151699328677,
      "loss": 5.2767,
      "mean_token_accuracy": 0.18176391273736953,
      "num_tokens": 18690847.0,
      "step": 10770
    },
    {
      "entropy": 5.641661310195923,
      "epoch": 0.8383582960513519,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004935085621755471,
      "loss": 5.2966,
      "mean_token_accuracy": 0.17963741570711136,
      "num_tokens": 18699658.0,
      "step": 10775
    },
    {
      "entropy": 5.578431844711304,
      "epoch": 0.8387473254230694,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493501951102707,
      "loss": 5.2742,
      "mean_token_accuracy": 0.1794950932264328,
      "num_tokens": 18708396.0,
      "step": 10780
    },
    {
      "entropy": 5.544351625442505,
      "epoch": 0.839136354794787,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004934953367144474,
      "loss": 5.2783,
      "mean_token_accuracy": 0.18206384629011155,
      "num_tokens": 18717185.0,
      "step": 10785
    },
    {
      "entropy": 5.63560905456543,
      "epoch": 0.8395253841665046,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004934887190108688,
      "loss": 5.231,
      "mean_token_accuracy": 0.18738207668066026,
      "num_tokens": 18725493.0,
      "step": 10790
    },
    {
      "entropy": 5.556916856765747,
      "epoch": 0.8399144135382222,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934820979920719,
      "loss": 5.2103,
      "mean_token_accuracy": 0.18609555065631866,
      "num_tokens": 18733843.0,
      "step": 10795
    },
    {
      "entropy": 5.542183446884155,
      "epoch": 0.8403034429099397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004934754736581567,
      "loss": 5.2885,
      "mean_token_accuracy": 0.17779247760772704,
      "num_tokens": 18742226.0,
      "step": 10800
    },
    {
      "entropy": 5.645714521408081,
      "epoch": 0.8406924722816572,
      "grad_norm": 1.0625,
      "learning_rate": 0.000493468846009224,
      "loss": 5.3957,
      "mean_token_accuracy": 0.177981199324131,
      "num_tokens": 18751117.0,
      "step": 10805
    },
    {
      "entropy": 5.56971435546875,
      "epoch": 0.8410815016533748,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004934622150453742,
      "loss": 5.2457,
      "mean_token_accuracy": 0.1807874247431755,
      "num_tokens": 18759244.0,
      "step": 10810
    },
    {
      "entropy": 5.562168884277344,
      "epoch": 0.8414705310250924,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493455580766708,
      "loss": 5.3012,
      "mean_token_accuracy": 0.17611841708421708,
      "num_tokens": 18768107.0,
      "step": 10815
    },
    {
      "entropy": 5.514830827713013,
      "epoch": 0.84185956039681,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934489431733262,
      "loss": 5.142,
      "mean_token_accuracy": 0.19034633338451384,
      "num_tokens": 18777419.0,
      "step": 10820
    },
    {
      "entropy": 5.557232809066773,
      "epoch": 0.8422485897685276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004934423022653293,
      "loss": 5.2635,
      "mean_token_accuracy": 0.18100958913564683,
      "num_tokens": 18786410.0,
      "step": 10825
    },
    {
      "entropy": 5.578015851974487,
      "epoch": 0.842637619140245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004934356580428182,
      "loss": 5.2297,
      "mean_token_accuracy": 0.1862280249595642,
      "num_tokens": 18794409.0,
      "step": 10830
    },
    {
      "entropy": 5.559137439727783,
      "epoch": 0.8430266485119626,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004934290105058937,
      "loss": 5.2897,
      "mean_token_accuracy": 0.17958732396364213,
      "num_tokens": 18803561.0,
      "step": 10835
    },
    {
      "entropy": 5.567430448532105,
      "epoch": 0.8434156778836802,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934223596546565,
      "loss": 5.2707,
      "mean_token_accuracy": 0.17618342638015747,
      "num_tokens": 18812096.0,
      "step": 10840
    },
    {
      "entropy": 5.5775820255279545,
      "epoch": 0.8438047072553978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004934157054892077,
      "loss": 5.2837,
      "mean_token_accuracy": 0.17901571393013,
      "num_tokens": 18820506.0,
      "step": 10845
    },
    {
      "entropy": 5.6024017333984375,
      "epoch": 0.8441937366271154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934090480096482,
      "loss": 5.2812,
      "mean_token_accuracy": 0.1773514449596405,
      "num_tokens": 18829437.0,
      "step": 10850
    },
    {
      "entropy": 5.66842794418335,
      "epoch": 0.8445827659988329,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004934023872160791,
      "loss": 5.3758,
      "mean_token_accuracy": 0.17594860941171647,
      "num_tokens": 18838759.0,
      "step": 10855
    },
    {
      "entropy": 5.5276306629180905,
      "epoch": 0.8449717953705504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004933957231086014,
      "loss": 5.2824,
      "mean_token_accuracy": 0.18068666160106658,
      "num_tokens": 18847000.0,
      "step": 10860
    },
    {
      "entropy": 5.623405265808105,
      "epoch": 0.845360824742268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933890556873161,
      "loss": 5.3551,
      "mean_token_accuracy": 0.17632968872785568,
      "num_tokens": 18855370.0,
      "step": 10865
    },
    {
      "entropy": 5.687759256362915,
      "epoch": 0.8457498541139856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004933823849523246,
      "loss": 5.4133,
      "mean_token_accuracy": 0.17654733955860138,
      "num_tokens": 18863825.0,
      "step": 10870
    },
    {
      "entropy": 5.577779293060303,
      "epoch": 0.8461388834857032,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004933757109037279,
      "loss": 5.1649,
      "mean_token_accuracy": 0.18300394713878632,
      "num_tokens": 18872088.0,
      "step": 10875
    },
    {
      "entropy": 5.532320165634156,
      "epoch": 0.8465279128574207,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004933690335416274,
      "loss": 5.1491,
      "mean_token_accuracy": 0.19310869574546813,
      "num_tokens": 18880472.0,
      "step": 10880
    },
    {
      "entropy": 5.597385549545288,
      "epoch": 0.8469169422291383,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933623528661245,
      "loss": 5.3645,
      "mean_token_accuracy": 0.17753133773803711,
      "num_tokens": 18889822.0,
      "step": 10885
    },
    {
      "entropy": 5.655607271194458,
      "epoch": 0.8473059716008559,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004933556688773203,
      "loss": 5.2962,
      "mean_token_accuracy": 0.17769573032855987,
      "num_tokens": 18897542.0,
      "step": 10890
    },
    {
      "entropy": 5.542865037918091,
      "epoch": 0.8476950009725734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933489815753165,
      "loss": 5.3324,
      "mean_token_accuracy": 0.17968873977661132,
      "num_tokens": 18906330.0,
      "step": 10895
    },
    {
      "entropy": 5.562378835678101,
      "epoch": 0.848084030344291,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004933422909602143,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1919698789715767,
      "num_tokens": 18914176.0,
      "step": 10900
    },
    {
      "entropy": 5.519755268096924,
      "epoch": 0.8484730597160085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933355970321155,
      "loss": 5.2027,
      "mean_token_accuracy": 0.19044993817806244,
      "num_tokens": 18922233.0,
      "step": 10905
    },
    {
      "entropy": 5.433318853378296,
      "epoch": 0.8488620890877261,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933288997911215,
      "loss": 5.2367,
      "mean_token_accuracy": 0.18599033653736113,
      "num_tokens": 18930742.0,
      "step": 10910
    },
    {
      "entropy": 5.564408254623413,
      "epoch": 0.8492511184594437,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933221992373338,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18545434772968292,
      "num_tokens": 18938879.0,
      "step": 10915
    },
    {
      "entropy": 5.717005825042724,
      "epoch": 0.8496401478311613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933154953708544,
      "loss": 5.3702,
      "mean_token_accuracy": 0.16363996118307114,
      "num_tokens": 18947753.0,
      "step": 10920
    },
    {
      "entropy": 5.589132356643677,
      "epoch": 0.8500291772028788,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933087881917848,
      "loss": 5.2418,
      "mean_token_accuracy": 0.17988715022802354,
      "num_tokens": 18956805.0,
      "step": 10925
    },
    {
      "entropy": 5.512399101257325,
      "epoch": 0.8504182065745964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933020777002268,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1815769702196121,
      "num_tokens": 18965683.0,
      "step": 10930
    },
    {
      "entropy": 5.5154742240905765,
      "epoch": 0.8508072359463139,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004932953638962823,
      "loss": 5.2143,
      "mean_token_accuracy": 0.18446161597967148,
      "num_tokens": 18973873.0,
      "step": 10935
    },
    {
      "entropy": 5.474437093734741,
      "epoch": 0.8511962653180315,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493288646780053,
      "loss": 5.1639,
      "mean_token_accuracy": 0.18976201564073564,
      "num_tokens": 18982097.0,
      "step": 10940
    },
    {
      "entropy": 5.501852941513062,
      "epoch": 0.8515852946897491,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004932819263516409,
      "loss": 5.2165,
      "mean_token_accuracy": 0.18093420565128326,
      "num_tokens": 18991019.0,
      "step": 10945
    },
    {
      "entropy": 5.600112009048462,
      "epoch": 0.8519743240614667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932752026111481,
      "loss": 5.2165,
      "mean_token_accuracy": 0.18231619596481324,
      "num_tokens": 18999231.0,
      "step": 10950
    },
    {
      "entropy": 5.6302252292633055,
      "epoch": 0.8523633534331843,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004932684755586765,
      "loss": 5.3031,
      "mean_token_accuracy": 0.18163848519325257,
      "num_tokens": 19007807.0,
      "step": 10955
    },
    {
      "entropy": 5.535033559799194,
      "epoch": 0.8527523828049017,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004932617451943282,
      "loss": 5.2752,
      "mean_token_accuracy": 0.1835549756884575,
      "num_tokens": 19017908.0,
      "step": 10960
    },
    {
      "entropy": 5.522093915939331,
      "epoch": 0.8531414121766193,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932550115182053,
      "loss": 5.2868,
      "mean_token_accuracy": 0.17447712868452073,
      "num_tokens": 19026999.0,
      "step": 10965
    },
    {
      "entropy": 5.500233602523804,
      "epoch": 0.8535304415483369,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049324827453041,
      "loss": 5.1884,
      "mean_token_accuracy": 0.19003772139549255,
      "num_tokens": 19035738.0,
      "step": 10970
    },
    {
      "entropy": 5.527360963821411,
      "epoch": 0.8539194709200545,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932415342310446,
      "loss": 5.2773,
      "mean_token_accuracy": 0.18088242411613464,
      "num_tokens": 19044530.0,
      "step": 10975
    },
    {
      "entropy": 5.537290048599243,
      "epoch": 0.8543085002917721,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004932347906202111,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18226035833358764,
      "num_tokens": 19053306.0,
      "step": 10980
    },
    {
      "entropy": 5.610524988174438,
      "epoch": 0.8546975296634896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493228043698012,
      "loss": 5.2643,
      "mean_token_accuracy": 0.17798605263233186,
      "num_tokens": 19062172.0,
      "step": 10985
    },
    {
      "entropy": 5.441015672683716,
      "epoch": 0.8550865590352071,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004932212934645498,
      "loss": 5.0494,
      "mean_token_accuracy": 0.1973587304353714,
      "num_tokens": 19070804.0,
      "step": 10990
    },
    {
      "entropy": 5.550571537017822,
      "epoch": 0.8554755884069247,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932145399199268,
      "loss": 5.2045,
      "mean_token_accuracy": 0.18815929740667342,
      "num_tokens": 19079055.0,
      "step": 10995
    },
    {
      "entropy": 5.519578647613526,
      "epoch": 0.8558646177786423,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932077830642455,
      "loss": 5.1932,
      "mean_token_accuracy": 0.18200305253267288,
      "num_tokens": 19087459.0,
      "step": 11000
    },
    {
      "entropy": 5.436576414108276,
      "epoch": 0.8562536471503599,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932010228976084,
      "loss": 5.0968,
      "mean_token_accuracy": 0.18955010622739793,
      "num_tokens": 19096271.0,
      "step": 11005
    },
    {
      "entropy": 5.591752910614014,
      "epoch": 0.8566426765220774,
      "grad_norm": 1.09375,
      "learning_rate": 0.000493194259420118,
      "loss": 5.2681,
      "mean_token_accuracy": 0.17966681718826294,
      "num_tokens": 19105257.0,
      "step": 11010
    },
    {
      "entropy": 5.586348581314087,
      "epoch": 0.857031705893795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000493187492631877,
      "loss": 5.2202,
      "mean_token_accuracy": 0.18629482686519622,
      "num_tokens": 19113673.0,
      "step": 11015
    },
    {
      "entropy": 5.532885360717773,
      "epoch": 0.8574207352655125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931807225329882,
      "loss": 5.2143,
      "mean_token_accuracy": 0.18833238929510115,
      "num_tokens": 19121635.0,
      "step": 11020
    },
    {
      "entropy": 5.533400058746338,
      "epoch": 0.8578097646372301,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931739491235541,
      "loss": 5.2485,
      "mean_token_accuracy": 0.18850345611572267,
      "num_tokens": 19130097.0,
      "step": 11025
    },
    {
      "entropy": 5.49743390083313,
      "epoch": 0.8581987940089477,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004931671724036777,
      "loss": 5.1609,
      "mean_token_accuracy": 0.18554106056690217,
      "num_tokens": 19138349.0,
      "step": 11030
    },
    {
      "entropy": 5.522438383102417,
      "epoch": 0.8585878233806652,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004931603923734616,
      "loss": 5.2654,
      "mean_token_accuracy": 0.18063793629407882,
      "num_tokens": 19146948.0,
      "step": 11035
    },
    {
      "entropy": 5.499011421203614,
      "epoch": 0.8589768527523828,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004931536090330088,
      "loss": 5.1995,
      "mean_token_accuracy": 0.19050881117582322,
      "num_tokens": 19155372.0,
      "step": 11040
    },
    {
      "entropy": 5.645586442947388,
      "epoch": 0.8593658821241004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004931468223824222,
      "loss": 5.3259,
      "mean_token_accuracy": 0.17474284917116165,
      "num_tokens": 19163634.0,
      "step": 11045
    },
    {
      "entropy": 5.665350770950317,
      "epoch": 0.859754911495818,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004931400324218048,
      "loss": 5.3209,
      "mean_token_accuracy": 0.1806083530187607,
      "num_tokens": 19171582.0,
      "step": 11050
    },
    {
      "entropy": 5.589558506011963,
      "epoch": 0.8601439408675355,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931332391512597,
      "loss": 5.3086,
      "mean_token_accuracy": 0.17354440689086914,
      "num_tokens": 19181136.0,
      "step": 11055
    },
    {
      "entropy": 5.591264867782593,
      "epoch": 0.860532970239253,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004931264425708897,
      "loss": 5.3011,
      "mean_token_accuracy": 0.17281284034252167,
      "num_tokens": 19189389.0,
      "step": 11060
    },
    {
      "entropy": 5.5910852432250975,
      "epoch": 0.8609219996109706,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004931196426807983,
      "loss": 5.2882,
      "mean_token_accuracy": 0.17940365821123122,
      "num_tokens": 19198315.0,
      "step": 11065
    },
    {
      "entropy": 5.651271152496338,
      "epoch": 0.8613110289826882,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004931128394810884,
      "loss": 5.3332,
      "mean_token_accuracy": 0.1748519644141197,
      "num_tokens": 19207234.0,
      "step": 11070
    },
    {
      "entropy": 5.581159782409668,
      "epoch": 0.8617000583544058,
      "grad_norm": 1.0,
      "learning_rate": 0.0004931060329718634,
      "loss": 5.3228,
      "mean_token_accuracy": 0.17085424363613128,
      "num_tokens": 19216552.0,
      "step": 11075
    },
    {
      "entropy": 5.557453680038452,
      "epoch": 0.8620890877261234,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004930992231532265,
      "loss": 5.2677,
      "mean_token_accuracy": 0.1751312792301178,
      "num_tokens": 19225237.0,
      "step": 11080
    },
    {
      "entropy": 5.515236663818359,
      "epoch": 0.8624781170978408,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493092410025281,
      "loss": 5.1836,
      "mean_token_accuracy": 0.19219851195812226,
      "num_tokens": 19233246.0,
      "step": 11085
    },
    {
      "entropy": 5.5010457038879395,
      "epoch": 0.8628671464695584,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930855935881302,
      "loss": 5.2127,
      "mean_token_accuracy": 0.18923863768577576,
      "num_tokens": 19241748.0,
      "step": 11090
    },
    {
      "entropy": 5.624505949020386,
      "epoch": 0.863256175841276,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930787738418777,
      "loss": 5.3639,
      "mean_token_accuracy": 0.17303334623575212,
      "num_tokens": 19250259.0,
      "step": 11095
    },
    {
      "entropy": 5.683621215820312,
      "epoch": 0.8636452052129936,
      "grad_norm": 1.0,
      "learning_rate": 0.0004930719507866269,
      "loss": 5.4206,
      "mean_token_accuracy": 0.1744420424103737,
      "num_tokens": 19259266.0,
      "step": 11100
    },
    {
      "entropy": 5.6021300792694095,
      "epoch": 0.8640342345847112,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004930651244224814,
      "loss": 5.2495,
      "mean_token_accuracy": 0.18053482621908187,
      "num_tokens": 19268099.0,
      "step": 11105
    },
    {
      "entropy": 5.544763803482056,
      "epoch": 0.8644232639564287,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930582947495448,
      "loss": 5.1575,
      "mean_token_accuracy": 0.1893578976392746,
      "num_tokens": 19276894.0,
      "step": 11110
    },
    {
      "entropy": 5.4840662479400635,
      "epoch": 0.8648122933281462,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004930514617679206,
      "loss": 5.2185,
      "mean_token_accuracy": 0.18783557116985322,
      "num_tokens": 19285724.0,
      "step": 11115
    },
    {
      "entropy": 5.5312365055084225,
      "epoch": 0.8652013226998638,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930446254777125,
      "loss": 5.1477,
      "mean_token_accuracy": 0.19058193862438202,
      "num_tokens": 19294468.0,
      "step": 11120
    },
    {
      "entropy": 5.5762135028839115,
      "epoch": 0.8655903520715814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004930377858790242,
      "loss": 5.2015,
      "mean_token_accuracy": 0.18814353048801422,
      "num_tokens": 19303575.0,
      "step": 11125
    },
    {
      "entropy": 5.6263727188110355,
      "epoch": 0.865979381443299,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004930309429719595,
      "loss": 5.3576,
      "mean_token_accuracy": 0.1824032887816429,
      "num_tokens": 19312778.0,
      "step": 11130
    },
    {
      "entropy": 5.603872346878052,
      "epoch": 0.8663684108150166,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004930240967566224,
      "loss": 5.2345,
      "mean_token_accuracy": 0.17890094667673112,
      "num_tokens": 19320725.0,
      "step": 11135
    },
    {
      "entropy": 5.565314292907715,
      "epoch": 0.8667574401867341,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930172472331167,
      "loss": 5.2472,
      "mean_token_accuracy": 0.18859666287899018,
      "num_tokens": 19329842.0,
      "step": 11140
    },
    {
      "entropy": 5.544046688079834,
      "epoch": 0.8671464695584516,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004930103944015463,
      "loss": 5.2522,
      "mean_token_accuracy": 0.18350773751735688,
      "num_tokens": 19338136.0,
      "step": 11145
    },
    {
      "entropy": 5.539253520965576,
      "epoch": 0.8675354989301692,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930035382620149,
      "loss": 5.218,
      "mean_token_accuracy": 0.18330558389425278,
      "num_tokens": 19346597.0,
      "step": 11150
    },
    {
      "entropy": 5.540123414993286,
      "epoch": 0.8679245283018868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492996678814627,
      "loss": 5.2079,
      "mean_token_accuracy": 0.18596190661191941,
      "num_tokens": 19355595.0,
      "step": 11155
    },
    {
      "entropy": 5.5605556011199955,
      "epoch": 0.8683135576736044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929898160594865,
      "loss": 5.279,
      "mean_token_accuracy": 0.1812897726893425,
      "num_tokens": 19364133.0,
      "step": 11160
    },
    {
      "entropy": 5.614044094085694,
      "epoch": 0.8687025870453219,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004929829499966974,
      "loss": 5.3125,
      "mean_token_accuracy": 0.17573751956224443,
      "num_tokens": 19373628.0,
      "step": 11165
    },
    {
      "entropy": 5.589564180374145,
      "epoch": 0.8690916164170395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004929760806263641,
      "loss": 5.2587,
      "mean_token_accuracy": 0.18214433491230012,
      "num_tokens": 19381901.0,
      "step": 11170
    },
    {
      "entropy": 5.577784156799316,
      "epoch": 0.8694806457887571,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004929692079485906,
      "loss": 5.3356,
      "mean_token_accuracy": 0.17620855271816255,
      "num_tokens": 19390390.0,
      "step": 11175
    },
    {
      "entropy": 5.590073919296264,
      "epoch": 0.8698696751604746,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929623319634814,
      "loss": 5.308,
      "mean_token_accuracy": 0.17319257110357283,
      "num_tokens": 19399858.0,
      "step": 11180
    },
    {
      "entropy": 5.546296644210815,
      "epoch": 0.8702587045321922,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929554526711407,
      "loss": 5.2542,
      "mean_token_accuracy": 0.17633968144655227,
      "num_tokens": 19409062.0,
      "step": 11185
    },
    {
      "entropy": 5.503489637374878,
      "epoch": 0.8706477339039097,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004929485700716729,
      "loss": 5.218,
      "mean_token_accuracy": 0.18555265814065933,
      "num_tokens": 19417228.0,
      "step": 11190
    },
    {
      "entropy": 5.571311902999878,
      "epoch": 0.8710367632756273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929416841651824,
      "loss": 5.3247,
      "mean_token_accuracy": 0.18117835223674775,
      "num_tokens": 19426086.0,
      "step": 11195
    },
    {
      "entropy": 5.61510443687439,
      "epoch": 0.8714257926473449,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929347949517739,
      "loss": 5.3117,
      "mean_token_accuracy": 0.17909523248672485,
      "num_tokens": 19435363.0,
      "step": 11200
    },
    {
      "entropy": 5.545312356948853,
      "epoch": 0.8718148220190625,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004929279024315516,
      "loss": 5.1139,
      "mean_token_accuracy": 0.18786819279193878,
      "num_tokens": 19443355.0,
      "step": 11205
    },
    {
      "entropy": 5.502433443069458,
      "epoch": 0.87220385139078,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929210066046204,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1818842500448227,
      "num_tokens": 19452217.0,
      "step": 11210
    },
    {
      "entropy": 5.608479261398315,
      "epoch": 0.8725928807624975,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004929141074710847,
      "loss": 5.3748,
      "mean_token_accuracy": 0.17233697474002838,
      "num_tokens": 19460034.0,
      "step": 11215
    },
    {
      "entropy": 5.5954773902893065,
      "epoch": 0.8729819101342151,
      "grad_norm": 1.125,
      "learning_rate": 0.0004929072050310491,
      "loss": 5.2977,
      "mean_token_accuracy": 0.18093376159667968,
      "num_tokens": 19468403.0,
      "step": 11220
    },
    {
      "entropy": 5.585196208953858,
      "epoch": 0.8733709395059327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929002992846188,
      "loss": 5.3213,
      "mean_token_accuracy": 0.17999297082424165,
      "num_tokens": 19477551.0,
      "step": 11225
    },
    {
      "entropy": 5.598291873931885,
      "epoch": 0.8737599688776503,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928933902318981,
      "loss": 5.2184,
      "mean_token_accuracy": 0.18642991483211518,
      "num_tokens": 19486131.0,
      "step": 11230
    },
    {
      "entropy": 5.61588864326477,
      "epoch": 0.8741489982493679,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000492886477872992,
      "loss": 5.3513,
      "mean_token_accuracy": 0.1688728526234627,
      "num_tokens": 19496212.0,
      "step": 11235
    },
    {
      "entropy": 5.56214451789856,
      "epoch": 0.8745380276210853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004928795622080054,
      "loss": 5.2042,
      "mean_token_accuracy": 0.1845085084438324,
      "num_tokens": 19504556.0,
      "step": 11240
    },
    {
      "entropy": 5.617821264266968,
      "epoch": 0.8749270569928029,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004928726432370434,
      "loss": 5.2609,
      "mean_token_accuracy": 0.1788235515356064,
      "num_tokens": 19512832.0,
      "step": 11245
    },
    {
      "entropy": 5.620820713043213,
      "epoch": 0.8753160863645205,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004928657209602107,
      "loss": 5.2655,
      "mean_token_accuracy": 0.18020354360342025,
      "num_tokens": 19521410.0,
      "step": 11250
    },
    {
      "entropy": 5.556787490844727,
      "epoch": 0.8757051157362381,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928587953776125,
      "loss": 5.2555,
      "mean_token_accuracy": 0.18206322193145752,
      "num_tokens": 19529935.0,
      "step": 11255
    },
    {
      "entropy": 5.491008043289185,
      "epoch": 0.8760941451079557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004928518664893538,
      "loss": 5.1554,
      "mean_token_accuracy": 0.186825954914093,
      "num_tokens": 19538330.0,
      "step": 11260
    },
    {
      "entropy": 5.522809934616089,
      "epoch": 0.8764831744796732,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928449342955397,
      "loss": 5.1491,
      "mean_token_accuracy": 0.18726811707019805,
      "num_tokens": 19546860.0,
      "step": 11265
    },
    {
      "entropy": 5.57469539642334,
      "epoch": 0.8768722038513908,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004928379987962756,
      "loss": 5.249,
      "mean_token_accuracy": 0.18167571127414703,
      "num_tokens": 19554945.0,
      "step": 11270
    },
    {
      "entropy": 5.599477624893188,
      "epoch": 0.8772612332231083,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928310599916666,
      "loss": 5.2726,
      "mean_token_accuracy": 0.18034662455320358,
      "num_tokens": 19563182.0,
      "step": 11275
    },
    {
      "entropy": 5.513549041748047,
      "epoch": 0.8776502625948259,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928241178818178,
      "loss": 5.2566,
      "mean_token_accuracy": 0.18222971856594086,
      "num_tokens": 19571405.0,
      "step": 11280
    },
    {
      "entropy": 5.65979118347168,
      "epoch": 0.8780392919665435,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004928171724668349,
      "loss": 5.3253,
      "mean_token_accuracy": 0.18145942240953444,
      "num_tokens": 19580053.0,
      "step": 11285
    },
    {
      "entropy": 5.6746344566345215,
      "epoch": 0.878428321338261,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928102237468229,
      "loss": 5.3502,
      "mean_token_accuracy": 0.17783455401659012,
      "num_tokens": 19588457.0,
      "step": 11290
    },
    {
      "entropy": 5.510228490829467,
      "epoch": 0.8788173507099786,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004928032717218876,
      "loss": 5.191,
      "mean_token_accuracy": 0.18567302823066711,
      "num_tokens": 19597608.0,
      "step": 11295
    },
    {
      "entropy": 5.530666351318359,
      "epoch": 0.8792063800816962,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004927963163921343,
      "loss": 5.2182,
      "mean_token_accuracy": 0.1877909481525421,
      "num_tokens": 19605904.0,
      "step": 11300
    },
    {
      "entropy": 5.585718584060669,
      "epoch": 0.8795954094534137,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004927893577576685,
      "loss": 5.2041,
      "mean_token_accuracy": 0.17550091743469237,
      "num_tokens": 19614660.0,
      "step": 11305
    },
    {
      "entropy": 5.533907270431518,
      "epoch": 0.8799844388251313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004927823958185959,
      "loss": 5.1828,
      "mean_token_accuracy": 0.19499629586935044,
      "num_tokens": 19623210.0,
      "step": 11310
    },
    {
      "entropy": 5.558951330184937,
      "epoch": 0.8803734681968488,
      "grad_norm": 0.96875,
      "learning_rate": 0.000492775430575022,
      "loss": 5.2722,
      "mean_token_accuracy": 0.17748644202947617,
      "num_tokens": 19631747.0,
      "step": 11315
    },
    {
      "entropy": 5.6451085090637205,
      "epoch": 0.8807624975685664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927684620270527,
      "loss": 5.3576,
      "mean_token_accuracy": 0.1731577232480049,
      "num_tokens": 19639902.0,
      "step": 11320
    },
    {
      "entropy": 5.6624321937561035,
      "epoch": 0.881151526940284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927614901747935,
      "loss": 5.3697,
      "mean_token_accuracy": 0.17460268884897232,
      "num_tokens": 19648478.0,
      "step": 11325
    },
    {
      "entropy": 5.690453720092774,
      "epoch": 0.8815405563120016,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004927545150183503,
      "loss": 5.389,
      "mean_token_accuracy": 0.17978238314390182,
      "num_tokens": 19657467.0,
      "step": 11330
    },
    {
      "entropy": 5.718032169342041,
      "epoch": 0.8819295856837192,
      "grad_norm": 1.109375,
      "learning_rate": 0.000492747536557829,
      "loss": 5.3469,
      "mean_token_accuracy": 0.16780038177967072,
      "num_tokens": 19665382.0,
      "step": 11335
    },
    {
      "entropy": 5.562109756469726,
      "epoch": 0.8823186150554367,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927405547933353,
      "loss": 5.1483,
      "mean_token_accuracy": 0.1870612844824791,
      "num_tokens": 19674486.0,
      "step": 11340
    },
    {
      "entropy": 5.597084808349609,
      "epoch": 0.8827076444271542,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004927335697249753,
      "loss": 5.1849,
      "mean_token_accuracy": 0.18209384381771088,
      "num_tokens": 19683060.0,
      "step": 11345
    },
    {
      "entropy": 5.4729695320129395,
      "epoch": 0.8830966737988718,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004927265813528549,
      "loss": 5.1536,
      "mean_token_accuracy": 0.19034045487642287,
      "num_tokens": 19691634.0,
      "step": 11350
    },
    {
      "entropy": 5.54167742729187,
      "epoch": 0.8834857031705894,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004927195896770804,
      "loss": 5.2583,
      "mean_token_accuracy": 0.17653674483299256,
      "num_tokens": 19699975.0,
      "step": 11355
    },
    {
      "entropy": 5.5695860385894775,
      "epoch": 0.883874732542307,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004927125946977574,
      "loss": 5.2123,
      "mean_token_accuracy": 0.18081354349851608,
      "num_tokens": 19708431.0,
      "step": 11360
    },
    {
      "entropy": 5.548543310165405,
      "epoch": 0.8842637619140246,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004927055964149923,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1743628427386284,
      "num_tokens": 19717189.0,
      "step": 11365
    },
    {
      "entropy": 5.59026985168457,
      "epoch": 0.884652791285742,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926985948288914,
      "loss": 5.2754,
      "mean_token_accuracy": 0.18580946773290635,
      "num_tokens": 19725855.0,
      "step": 11370
    },
    {
      "entropy": 5.547528028488159,
      "epoch": 0.8850418206574596,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926915899395608,
      "loss": 5.204,
      "mean_token_accuracy": 0.18512689471244811,
      "num_tokens": 19734617.0,
      "step": 11375
    },
    {
      "entropy": 5.642363119125366,
      "epoch": 0.8854308500291772,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004926845817471068,
      "loss": 5.3033,
      "mean_token_accuracy": 0.17751339375972747,
      "num_tokens": 19743731.0,
      "step": 11380
    },
    {
      "entropy": 5.5991583347320555,
      "epoch": 0.8858198794008948,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004926775702516358,
      "loss": 5.2551,
      "mean_token_accuracy": 0.18061378598213196,
      "num_tokens": 19752336.0,
      "step": 11385
    },
    {
      "entropy": 5.555872297286987,
      "epoch": 0.8862089087726124,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004926705554532541,
      "loss": 5.2787,
      "mean_token_accuracy": 0.1789202481508255,
      "num_tokens": 19761488.0,
      "step": 11390
    },
    {
      "entropy": 5.5467212200164795,
      "epoch": 0.8865979381443299,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004926635373520682,
      "loss": 5.1844,
      "mean_token_accuracy": 0.18448866307735443,
      "num_tokens": 19770330.0,
      "step": 11395
    },
    {
      "entropy": 5.56444525718689,
      "epoch": 0.8869869675160474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926565159481845,
      "loss": 5.2244,
      "mean_token_accuracy": 0.18237495124340058,
      "num_tokens": 19779037.0,
      "step": 11400
    },
    {
      "entropy": 5.583184480667114,
      "epoch": 0.887375996887765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004926494912417097,
      "loss": 5.3047,
      "mean_token_accuracy": 0.17376061528921127,
      "num_tokens": 19787661.0,
      "step": 11405
    },
    {
      "entropy": 5.556111907958984,
      "epoch": 0.8877650262594826,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926424632327503,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18178965747356415,
      "num_tokens": 19796421.0,
      "step": 11410
    },
    {
      "entropy": 5.535477924346924,
      "epoch": 0.8881540556312002,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004926354319214129,
      "loss": 5.2206,
      "mean_token_accuracy": 0.18311661928892137,
      "num_tokens": 19805842.0,
      "step": 11415
    },
    {
      "entropy": 5.5428627014160154,
      "epoch": 0.8885430850029177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926283973078041,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18432259261608125,
      "num_tokens": 19814179.0,
      "step": 11420
    },
    {
      "entropy": 5.540349626541138,
      "epoch": 0.8889321143746353,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926213593920309,
      "loss": 5.3165,
      "mean_token_accuracy": 0.17773068845272064,
      "num_tokens": 19822435.0,
      "step": 11425
    },
    {
      "entropy": 5.614234161376953,
      "epoch": 0.8893211437463528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004926143181742,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18455283790826799,
      "num_tokens": 19830300.0,
      "step": 11430
    },
    {
      "entropy": 5.556007957458496,
      "epoch": 0.8897101731180704,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926072736544181,
      "loss": 5.3444,
      "mean_token_accuracy": 0.17781427651643752,
      "num_tokens": 19839264.0,
      "step": 11435
    },
    {
      "entropy": 5.5439025402069095,
      "epoch": 0.890099202489788,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926002258327922,
      "loss": 5.2694,
      "mean_token_accuracy": 0.18135368525981904,
      "num_tokens": 19848086.0,
      "step": 11440
    },
    {
      "entropy": 5.610151863098144,
      "epoch": 0.8904882318615055,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004925931747094292,
      "loss": 5.2579,
      "mean_token_accuracy": 0.1756952151656151,
      "num_tokens": 19856571.0,
      "step": 11445
    },
    {
      "entropy": 5.570679187774658,
      "epoch": 0.8908772612332231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004925861202844361,
      "loss": 5.277,
      "mean_token_accuracy": 0.1755858764052391,
      "num_tokens": 19865272.0,
      "step": 11450
    },
    {
      "entropy": 5.539387464523315,
      "epoch": 0.8912662906049407,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049257906255792,
      "loss": 5.1969,
      "mean_token_accuracy": 0.1824360743165016,
      "num_tokens": 19874307.0,
      "step": 11455
    },
    {
      "entropy": 5.558570623397827,
      "epoch": 0.8916553199766583,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492572001529988,
      "loss": 5.2803,
      "mean_token_accuracy": 0.17733534574508666,
      "num_tokens": 19882851.0,
      "step": 11460
    },
    {
      "entropy": 5.556431293487549,
      "epoch": 0.8920443493483758,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925649372007469,
      "loss": 5.1565,
      "mean_token_accuracy": 0.19295447170734406,
      "num_tokens": 19891276.0,
      "step": 11465
    },
    {
      "entropy": 5.616119909286499,
      "epoch": 0.8924333787200933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925578695703045,
      "loss": 5.3134,
      "mean_token_accuracy": 0.1715076133608818,
      "num_tokens": 19900551.0,
      "step": 11470
    },
    {
      "entropy": 5.639095497131348,
      "epoch": 0.8928224080918109,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925507986387676,
      "loss": 5.33,
      "mean_token_accuracy": 0.17884072959423064,
      "num_tokens": 19909412.0,
      "step": 11475
    },
    {
      "entropy": 5.569097280502319,
      "epoch": 0.8932114374635285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004925437244062436,
      "loss": 5.1774,
      "mean_token_accuracy": 0.18880049884319305,
      "num_tokens": 19918073.0,
      "step": 11480
    },
    {
      "entropy": 5.445699977874756,
      "epoch": 0.8936004668352461,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004925366468728397,
      "loss": 5.134,
      "mean_token_accuracy": 0.19442103207111358,
      "num_tokens": 19926040.0,
      "step": 11485
    },
    {
      "entropy": 5.560358476638794,
      "epoch": 0.8939894962069637,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925295660386635,
      "loss": 5.2952,
      "mean_token_accuracy": 0.18159547597169876,
      "num_tokens": 19934543.0,
      "step": 11490
    },
    {
      "entropy": 5.584822463989258,
      "epoch": 0.8943785255786811,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004925224819038224,
      "loss": 5.2261,
      "mean_token_accuracy": 0.1877506196498871,
      "num_tokens": 19943198.0,
      "step": 11495
    },
    {
      "entropy": 5.501182699203492,
      "epoch": 0.8947675549503987,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925153944684239,
      "loss": 5.1606,
      "mean_token_accuracy": 0.19049780517816545,
      "num_tokens": 19952319.0,
      "step": 11500
    },
    {
      "entropy": 5.500726127624512,
      "epoch": 0.8951565843221163,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925083037325754,
      "loss": 5.1564,
      "mean_token_accuracy": 0.18824374675750732,
      "num_tokens": 19961982.0,
      "step": 11505
    },
    {
      "entropy": 5.541096115112305,
      "epoch": 0.8955456136938339,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004925012096963847,
      "loss": 5.1747,
      "mean_token_accuracy": 0.18422329127788545,
      "num_tokens": 19971055.0,
      "step": 11510
    },
    {
      "entropy": 5.46767144203186,
      "epoch": 0.8959346430655515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924941123599593,
      "loss": 5.0646,
      "mean_token_accuracy": 0.1973741590976715,
      "num_tokens": 19979850.0,
      "step": 11515
    },
    {
      "entropy": 5.563228511810303,
      "epoch": 0.896323672437269,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004924870117234069,
      "loss": 5.2763,
      "mean_token_accuracy": 0.18018045723438264,
      "num_tokens": 19989261.0,
      "step": 11520
    },
    {
      "entropy": 5.5026203155517575,
      "epoch": 0.8967127018089865,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004924799077868353,
      "loss": 5.1871,
      "mean_token_accuracy": 0.17769325226545335,
      "num_tokens": 19998321.0,
      "step": 11525
    },
    {
      "entropy": 5.500790166854858,
      "epoch": 0.8971017311807041,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004924728005503522,
      "loss": 5.2668,
      "mean_token_accuracy": 0.17981211543083192,
      "num_tokens": 20007191.0,
      "step": 11530
    },
    {
      "entropy": 5.615214252471924,
      "epoch": 0.8974907605524217,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924656900140655,
      "loss": 5.2881,
      "mean_token_accuracy": 0.17898485511541368,
      "num_tokens": 20015805.0,
      "step": 11535
    },
    {
      "entropy": 5.587937355041504,
      "epoch": 0.8978797899241393,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004924585761780831,
      "loss": 5.2651,
      "mean_token_accuracy": 0.1849955752491951,
      "num_tokens": 20024861.0,
      "step": 11540
    },
    {
      "entropy": 5.652350950241089,
      "epoch": 0.8982688192958569,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492451459042513,
      "loss": 5.3174,
      "mean_token_accuracy": 0.17797672301530837,
      "num_tokens": 20034208.0,
      "step": 11545
    },
    {
      "entropy": 5.544576501846313,
      "epoch": 0.8986578486675744,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492444338607463,
      "loss": 5.2727,
      "mean_token_accuracy": 0.1784976452589035,
      "num_tokens": 20042786.0,
      "step": 11550
    },
    {
      "entropy": 5.639313840866089,
      "epoch": 0.899046878039292,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004924372148730413,
      "loss": 5.322,
      "mean_token_accuracy": 0.17717861235141755,
      "num_tokens": 20051697.0,
      "step": 11555
    },
    {
      "entropy": 5.622322702407837,
      "epoch": 0.8994359074110095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492430087839356,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18705434650182723,
      "num_tokens": 20059814.0,
      "step": 11560
    },
    {
      "entropy": 5.555255651473999,
      "epoch": 0.8998249367827271,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924229575065152,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18103350102901458,
      "num_tokens": 20068329.0,
      "step": 11565
    },
    {
      "entropy": 5.581609153747559,
      "epoch": 0.9002139661544447,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492415823874627,
      "loss": 5.2698,
      "mean_token_accuracy": 0.18550906479358673,
      "num_tokens": 20076925.0,
      "step": 11570
    },
    {
      "entropy": 5.56740369796753,
      "epoch": 0.9006029955261622,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004924086869437998,
      "loss": 5.2909,
      "mean_token_accuracy": 0.17394916862249374,
      "num_tokens": 20085248.0,
      "step": 11575
    },
    {
      "entropy": 5.590228033065796,
      "epoch": 0.9009920248978798,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004924015467141417,
      "loss": 5.2863,
      "mean_token_accuracy": 0.17664130926132202,
      "num_tokens": 20093929.0,
      "step": 11580
    },
    {
      "entropy": 5.58238582611084,
      "epoch": 0.9013810542695974,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004923944031857613,
      "loss": 5.3472,
      "mean_token_accuracy": 0.17677741944789888,
      "num_tokens": 20103141.0,
      "step": 11585
    },
    {
      "entropy": 5.54950180053711,
      "epoch": 0.901770083641315,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923872563587668,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19261873364448548,
      "num_tokens": 20111938.0,
      "step": 11590
    },
    {
      "entropy": 5.52837119102478,
      "epoch": 0.9021591130130325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004923801062332666,
      "loss": 5.167,
      "mean_token_accuracy": 0.1935260236263275,
      "num_tokens": 20120795.0,
      "step": 11595
    },
    {
      "entropy": 5.56316065788269,
      "epoch": 0.90254814238475,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004923729528093694,
      "loss": 5.2534,
      "mean_token_accuracy": 0.18112141489982606,
      "num_tokens": 20129816.0,
      "step": 11600
    },
    {
      "entropy": 5.610924768447876,
      "epoch": 0.9029371717564676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004923657960871836,
      "loss": 5.38,
      "mean_token_accuracy": 0.17450439035892487,
      "num_tokens": 20138380.0,
      "step": 11605
    },
    {
      "entropy": 5.656967687606811,
      "epoch": 0.9033262011281852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923586360668178,
      "loss": 5.3778,
      "mean_token_accuracy": 0.1771547645330429,
      "num_tokens": 20147201.0,
      "step": 11610
    },
    {
      "entropy": 5.56103835105896,
      "epoch": 0.9037152304999028,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004923514727483807,
      "loss": 5.1967,
      "mean_token_accuracy": 0.18787529468536376,
      "num_tokens": 20156744.0,
      "step": 11615
    },
    {
      "entropy": 5.582573461532593,
      "epoch": 0.9041042598716204,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000492344306131981,
      "loss": 5.2335,
      "mean_token_accuracy": 0.17903553545475007,
      "num_tokens": 20165081.0,
      "step": 11620
    },
    {
      "entropy": 5.592436456680298,
      "epoch": 0.9044932892433378,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923371362177272,
      "loss": 5.2368,
      "mean_token_accuracy": 0.17818743586540223,
      "num_tokens": 20172893.0,
      "step": 11625
    },
    {
      "entropy": 5.575743532180786,
      "epoch": 0.9048823186150554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923299630057284,
      "loss": 5.2937,
      "mean_token_accuracy": 0.17944218814373017,
      "num_tokens": 20181351.0,
      "step": 11630
    },
    {
      "entropy": 5.594365119934082,
      "epoch": 0.905271347986773,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004923227864960934,
      "loss": 5.3206,
      "mean_token_accuracy": 0.17152225077152253,
      "num_tokens": 20189650.0,
      "step": 11635
    },
    {
      "entropy": 5.6272036075592045,
      "epoch": 0.9056603773584906,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492315606688931,
      "loss": 5.2667,
      "mean_token_accuracy": 0.17608900666236876,
      "num_tokens": 20198184.0,
      "step": 11640
    },
    {
      "entropy": 5.6504967212677,
      "epoch": 0.9060494067302082,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004923084235843501,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1882951557636261,
      "num_tokens": 20206357.0,
      "step": 11645
    },
    {
      "entropy": 5.51787371635437,
      "epoch": 0.9064384361019256,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923012371824598,
      "loss": 5.3138,
      "mean_token_accuracy": 0.1831569939851761,
      "num_tokens": 20214384.0,
      "step": 11650
    },
    {
      "entropy": 5.569528341293335,
      "epoch": 0.9068274654736432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492294047483369,
      "loss": 5.2594,
      "mean_token_accuracy": 0.17942496240139008,
      "num_tokens": 20224103.0,
      "step": 11655
    },
    {
      "entropy": 5.611847877502441,
      "epoch": 0.9072164948453608,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004922868544871869,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18178559690713883,
      "num_tokens": 20233414.0,
      "step": 11660
    },
    {
      "entropy": 5.593952655792236,
      "epoch": 0.9076055242170784,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922796581940227,
      "loss": 5.1803,
      "mean_token_accuracy": 0.19224899709224702,
      "num_tokens": 20240988.0,
      "step": 11665
    },
    {
      "entropy": 5.481856918334961,
      "epoch": 0.907994553588796,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922724586039855,
      "loss": 5.1899,
      "mean_token_accuracy": 0.18444481790065764,
      "num_tokens": 20250066.0,
      "step": 11670
    },
    {
      "entropy": 5.5165046691894535,
      "epoch": 0.9083835829605135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922652557171846,
      "loss": 5.1441,
      "mean_token_accuracy": 0.18613797426223755,
      "num_tokens": 20258435.0,
      "step": 11675
    },
    {
      "entropy": 5.574391174316406,
      "epoch": 0.9087726123322311,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922580495337292,
      "loss": 5.2908,
      "mean_token_accuracy": 0.18753574043512344,
      "num_tokens": 20267345.0,
      "step": 11680
    },
    {
      "entropy": 5.622555875778199,
      "epoch": 0.9091616417039486,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922508400537288,
      "loss": 5.3052,
      "mean_token_accuracy": 0.18619502782821656,
      "num_tokens": 20275644.0,
      "step": 11685
    },
    {
      "entropy": 5.602553844451904,
      "epoch": 0.9095506710756662,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922436272772926,
      "loss": 5.3039,
      "mean_token_accuracy": 0.18190854787826538,
      "num_tokens": 20284459.0,
      "step": 11690
    },
    {
      "entropy": 5.574173355102539,
      "epoch": 0.9099397004473838,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004922364112045301,
      "loss": 5.1817,
      "mean_token_accuracy": 0.18159767091274262,
      "num_tokens": 20293088.0,
      "step": 11695
    },
    {
      "entropy": 5.509349203109741,
      "epoch": 0.9103287298191013,
      "grad_norm": 1.09375,
      "learning_rate": 0.000492229191835551,
      "loss": 5.216,
      "mean_token_accuracy": 0.18721252381801606,
      "num_tokens": 20301811.0,
      "step": 11700
    },
    {
      "entropy": 5.6459808349609375,
      "epoch": 0.9107177591908189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004922219691704645,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1744689628481865,
      "num_tokens": 20310365.0,
      "step": 11705
    },
    {
      "entropy": 5.652016544342041,
      "epoch": 0.9111067885625365,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004922147432093805,
      "loss": 5.2731,
      "mean_token_accuracy": 0.18861508071422578,
      "num_tokens": 20318891.0,
      "step": 11710
    },
    {
      "entropy": 5.527889442443848,
      "epoch": 0.911495817934254,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922075139524083,
      "loss": 5.1479,
      "mean_token_accuracy": 0.19287789911031722,
      "num_tokens": 20327313.0,
      "step": 11715
    },
    {
      "entropy": 5.514367198944091,
      "epoch": 0.9118848473059716,
      "grad_norm": 1.078125,
      "learning_rate": 0.000492200281399658,
      "loss": 5.1379,
      "mean_token_accuracy": 0.1912658095359802,
      "num_tokens": 20335511.0,
      "step": 11720
    },
    {
      "entropy": 5.530355930328369,
      "epoch": 0.9122738766776891,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492193045551239,
      "loss": 5.1654,
      "mean_token_accuracy": 0.19051099866628646,
      "num_tokens": 20343588.0,
      "step": 11725
    },
    {
      "entropy": 5.557232236862182,
      "epoch": 0.9126629060494067,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921858064072612,
      "loss": 5.2391,
      "mean_token_accuracy": 0.1865159258246422,
      "num_tokens": 20352793.0,
      "step": 11730
    },
    {
      "entropy": 5.573578262329102,
      "epoch": 0.9130519354211243,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004921785639678347,
      "loss": 5.3133,
      "mean_token_accuracy": 0.182097165286541,
      "num_tokens": 20362480.0,
      "step": 11735
    },
    {
      "entropy": 5.5420506477355955,
      "epoch": 0.9134409647928419,
      "grad_norm": 1.015625,
      "learning_rate": 0.000492171318233069,
      "loss": 5.1714,
      "mean_token_accuracy": 0.18888458162546157,
      "num_tokens": 20371192.0,
      "step": 11740
    },
    {
      "entropy": 5.638186502456665,
      "epoch": 0.9138299941645595,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004921640692030742,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18450805842876433,
      "num_tokens": 20379322.0,
      "step": 11745
    },
    {
      "entropy": 5.622485828399658,
      "epoch": 0.914219023536277,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921568168779603,
      "loss": 5.2604,
      "mean_token_accuracy": 0.17219553589820863,
      "num_tokens": 20387976.0,
      "step": 11750
    },
    {
      "entropy": 5.5516157150268555,
      "epoch": 0.9146080529079945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004921495612578374,
      "loss": 5.325,
      "mean_token_accuracy": 0.17372918725013733,
      "num_tokens": 20397398.0,
      "step": 11755
    },
    {
      "entropy": 5.529106760025025,
      "epoch": 0.9149970822797121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004921423023428156,
      "loss": 5.2339,
      "mean_token_accuracy": 0.1838798388838768,
      "num_tokens": 20406503.0,
      "step": 11760
    },
    {
      "entropy": 5.541811323165893,
      "epoch": 0.9153861116514297,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004921350401330049,
      "loss": 5.1691,
      "mean_token_accuracy": 0.18593801259994508,
      "num_tokens": 20415117.0,
      "step": 11765
    },
    {
      "entropy": 5.5522106170654295,
      "epoch": 0.9157751410231473,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004921277746285155,
      "loss": 5.254,
      "mean_token_accuracy": 0.18175882548093797,
      "num_tokens": 20423519.0,
      "step": 11770
    },
    {
      "entropy": 5.561598682403565,
      "epoch": 0.9161641703948649,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004921205058294579,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1842939481139183,
      "num_tokens": 20431831.0,
      "step": 11775
    },
    {
      "entropy": 5.619694757461548,
      "epoch": 0.9165531997665823,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492113233735942,
      "loss": 5.3216,
      "mean_token_accuracy": 0.17371482849121095,
      "num_tokens": 20440927.0,
      "step": 11780
    },
    {
      "entropy": 5.524583244323731,
      "epoch": 0.9169422291382999,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921059583480785,
      "loss": 5.1934,
      "mean_token_accuracy": 0.19074779599905015,
      "num_tokens": 20448974.0,
      "step": 11785
    },
    {
      "entropy": 5.50584135055542,
      "epoch": 0.9173312585100175,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004920986796659775,
      "loss": 5.1245,
      "mean_token_accuracy": 0.191385318338871,
      "num_tokens": 20457955.0,
      "step": 11790
    },
    {
      "entropy": 5.507561826705933,
      "epoch": 0.9177202878817351,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004920913976897498,
      "loss": 5.2366,
      "mean_token_accuracy": 0.18471247851848602,
      "num_tokens": 20465942.0,
      "step": 11795
    },
    {
      "entropy": 5.541639804840088,
      "epoch": 0.9181093172534527,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004920841124195055,
      "loss": 5.2782,
      "mean_token_accuracy": 0.1835801288485527,
      "num_tokens": 20474726.0,
      "step": 11800
    },
    {
      "entropy": 5.513240098953247,
      "epoch": 0.9184983466251702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004920768238553554,
      "loss": 5.225,
      "mean_token_accuracy": 0.184701868891716,
      "num_tokens": 20483442.0,
      "step": 11805
    },
    {
      "entropy": 5.509365367889404,
      "epoch": 0.9188873759968877,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920695319974099,
      "loss": 5.1861,
      "mean_token_accuracy": 0.18636468350887297,
      "num_tokens": 20492950.0,
      "step": 11810
    },
    {
      "entropy": 5.607105827331543,
      "epoch": 0.9192764053686053,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920622368457798,
      "loss": 5.2863,
      "mean_token_accuracy": 0.1794748067855835,
      "num_tokens": 20502035.0,
      "step": 11815
    },
    {
      "entropy": 5.601960372924805,
      "epoch": 0.9196654347403229,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920549384005759,
      "loss": 5.1661,
      "mean_token_accuracy": 0.19141081124544143,
      "num_tokens": 20510140.0,
      "step": 11820
    },
    {
      "entropy": 5.587213468551636,
      "epoch": 0.9200544641120405,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920476366619086,
      "loss": 5.2799,
      "mean_token_accuracy": 0.17743276059627533,
      "num_tokens": 20519391.0,
      "step": 11825
    },
    {
      "entropy": 5.6275735855102536,
      "epoch": 0.920443493483758,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000492040331629889,
      "loss": 5.2717,
      "mean_token_accuracy": 0.1765116959810257,
      "num_tokens": 20527635.0,
      "step": 11830
    },
    {
      "entropy": 5.6019606590271,
      "epoch": 0.9208325228554756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920330233046277,
      "loss": 5.2326,
      "mean_token_accuracy": 0.18538276851177216,
      "num_tokens": 20536455.0,
      "step": 11835
    },
    {
      "entropy": 5.637334966659546,
      "epoch": 0.9212215522271932,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004920257116862359,
      "loss": 5.3601,
      "mean_token_accuracy": 0.17379544377326966,
      "num_tokens": 20545887.0,
      "step": 11840
    },
    {
      "entropy": 5.527892971038819,
      "epoch": 0.9216105815989107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004920183967748241,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18469038903713225,
      "num_tokens": 20554415.0,
      "step": 11845
    },
    {
      "entropy": 5.608368301391602,
      "epoch": 0.9219996109706283,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004920110785705037,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1792415663599968,
      "num_tokens": 20563081.0,
      "step": 11850
    },
    {
      "entropy": 5.5365087509155275,
      "epoch": 0.9223886403423458,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004920037570733857,
      "loss": 5.1588,
      "mean_token_accuracy": 0.1892438307404518,
      "num_tokens": 20571800.0,
      "step": 11855
    },
    {
      "entropy": 5.579374027252197,
      "epoch": 0.9227776697140634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919964322835809,
      "loss": 5.2944,
      "mean_token_accuracy": 0.18046103715896605,
      "num_tokens": 20580481.0,
      "step": 11860
    },
    {
      "entropy": 5.530831289291382,
      "epoch": 0.923166699085781,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004919891042012006,
      "loss": 5.1701,
      "mean_token_accuracy": 0.18944658488035201,
      "num_tokens": 20589575.0,
      "step": 11865
    },
    {
      "entropy": 5.531797075271607,
      "epoch": 0.9235557284574986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004919817728263562,
      "loss": 5.2465,
      "mean_token_accuracy": 0.18160925060510635,
      "num_tokens": 20598277.0,
      "step": 11870
    },
    {
      "entropy": 5.634156799316406,
      "epoch": 0.9239447578292161,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004919744381591586,
      "loss": 5.3722,
      "mean_token_accuracy": 0.18001783341169358,
      "num_tokens": 20607013.0,
      "step": 11875
    },
    {
      "entropy": 5.574320554733276,
      "epoch": 0.9243337872009336,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919671001997193,
      "loss": 5.278,
      "mean_token_accuracy": 0.17825234234333037,
      "num_tokens": 20615605.0,
      "step": 11880
    },
    {
      "entropy": 5.6079527854919435,
      "epoch": 0.9247228165726512,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919597589481497,
      "loss": 5.2612,
      "mean_token_accuracy": 0.1799523875117302,
      "num_tokens": 20624275.0,
      "step": 11885
    },
    {
      "entropy": 5.57183518409729,
      "epoch": 0.9251118459443688,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000491952414404561,
      "loss": 5.2162,
      "mean_token_accuracy": 0.1849514827132225,
      "num_tokens": 20632681.0,
      "step": 11890
    },
    {
      "entropy": 5.66807599067688,
      "epoch": 0.9255008753160864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919450665690646,
      "loss": 5.2688,
      "mean_token_accuracy": 0.18525569140911102,
      "num_tokens": 20641432.0,
      "step": 11895
    },
    {
      "entropy": 5.583403253555298,
      "epoch": 0.925889904687804,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004919377154417724,
      "loss": 5.2715,
      "mean_token_accuracy": 0.18098506182432175,
      "num_tokens": 20650120.0,
      "step": 11900
    },
    {
      "entropy": 5.529873466491699,
      "epoch": 0.9262789340595214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004919303610227954,
      "loss": 5.1948,
      "mean_token_accuracy": 0.18440381437540054,
      "num_tokens": 20658188.0,
      "step": 11905
    },
    {
      "entropy": 5.492690134048462,
      "epoch": 0.926667963431239,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919230033122457,
      "loss": 5.1105,
      "mean_token_accuracy": 0.19484479129314422,
      "num_tokens": 20666157.0,
      "step": 11910
    },
    {
      "entropy": 5.544192504882813,
      "epoch": 0.9270569928029566,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004919156423102345,
      "loss": 5.3112,
      "mean_token_accuracy": 0.18255216404795646,
      "num_tokens": 20674995.0,
      "step": 11915
    },
    {
      "entropy": 5.630472421646118,
      "epoch": 0.9274460221746742,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004919082780168736,
      "loss": 5.2452,
      "mean_token_accuracy": 0.17868777960538865,
      "num_tokens": 20683786.0,
      "step": 11920
    },
    {
      "entropy": 5.580710172653198,
      "epoch": 0.9278350515463918,
      "grad_norm": 1.0,
      "learning_rate": 0.0004919009104322751,
      "loss": 5.2721,
      "mean_token_accuracy": 0.17575779557228088,
      "num_tokens": 20692582.0,
      "step": 11925
    },
    {
      "entropy": 5.596516084671021,
      "epoch": 0.9282240809181093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004918935395565503,
      "loss": 5.32,
      "mean_token_accuracy": 0.17592122703790664,
      "num_tokens": 20700892.0,
      "step": 11930
    },
    {
      "entropy": 5.51577353477478,
      "epoch": 0.9286131102898268,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004918861653898113,
      "loss": 5.137,
      "mean_token_accuracy": 0.196639546751976,
      "num_tokens": 20709521.0,
      "step": 11935
    },
    {
      "entropy": 5.532233953475952,
      "epoch": 0.9290021396615444,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004918787879321701,
      "loss": 5.2632,
      "mean_token_accuracy": 0.17572422772645951,
      "num_tokens": 20717614.0,
      "step": 11940
    },
    {
      "entropy": 5.560285663604736,
      "epoch": 0.929391169033262,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918714071837384,
      "loss": 5.1965,
      "mean_token_accuracy": 0.18053747564554215,
      "num_tokens": 20726624.0,
      "step": 11945
    },
    {
      "entropy": 5.6247334480285645,
      "epoch": 0.9297801984049796,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004918640231446282,
      "loss": 5.2943,
      "mean_token_accuracy": 0.1780096933245659,
      "num_tokens": 20734738.0,
      "step": 11950
    },
    {
      "entropy": 5.582316732406616,
      "epoch": 0.9301692277766972,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004918566358149517,
      "loss": 5.2306,
      "mean_token_accuracy": 0.1833841785788536,
      "num_tokens": 20742825.0,
      "step": 11955
    },
    {
      "entropy": 5.552599716186523,
      "epoch": 0.9305582571484147,
      "grad_norm": 0.9375,
      "learning_rate": 0.000491849245194821,
      "loss": 5.1919,
      "mean_token_accuracy": 0.1891092151403427,
      "num_tokens": 20752132.0,
      "step": 11960
    },
    {
      "entropy": 5.628361988067627,
      "epoch": 0.9309472865201323,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918418512843479,
      "loss": 5.3183,
      "mean_token_accuracy": 0.1766253247857094,
      "num_tokens": 20761838.0,
      "step": 11965
    },
    {
      "entropy": 5.599145078659058,
      "epoch": 0.9313363158918498,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918344540836451,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1821851596236229,
      "num_tokens": 20770405.0,
      "step": 11970
    },
    {
      "entropy": 5.711616325378418,
      "epoch": 0.9317253452635674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004918270535928244,
      "loss": 5.3912,
      "mean_token_accuracy": 0.17361809313297272,
      "num_tokens": 20780130.0,
      "step": 11975
    },
    {
      "entropy": 5.53233323097229,
      "epoch": 0.932114374635285,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004918196498119984,
      "loss": 5.2735,
      "mean_token_accuracy": 0.18742605298757553,
      "num_tokens": 20788852.0,
      "step": 11980
    },
    {
      "entropy": 5.545420932769775,
      "epoch": 0.9325034040070025,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918122427412793,
      "loss": 5.2168,
      "mean_token_accuracy": 0.18332792818546295,
      "num_tokens": 20796944.0,
      "step": 11985
    },
    {
      "entropy": 5.549014568328857,
      "epoch": 0.9328924333787201,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918048323807795,
      "loss": 5.3128,
      "mean_token_accuracy": 0.18216352313756942,
      "num_tokens": 20806118.0,
      "step": 11990
    },
    {
      "entropy": 5.609514474868774,
      "epoch": 0.9332814627504377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917974187306114,
      "loss": 5.1055,
      "mean_token_accuracy": 0.1885119527578354,
      "num_tokens": 20814643.0,
      "step": 11995
    },
    {
      "entropy": 5.583715391159058,
      "epoch": 0.9336704921221552,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917900017908875,
      "loss": 5.1952,
      "mean_token_accuracy": 0.18639011681079865,
      "num_tokens": 20823315.0,
      "step": 12000
    },
    {
      "epoch": 0.9336704921221552,
      "eval_entropy": 5.316801309631465,
      "eval_loss": 5.258728981018066,
      "eval_mean_token_accuracy": 0.19031005386725655,
      "eval_num_tokens": 20823315.0,
      "eval_runtime": 21.8591,
      "eval_samples_per_second": 1901.173,
      "eval_steps_per_second": 237.658,
      "step": 12000
    },
    {
      "entropy": 5.508055639266968,
      "epoch": 0.9340595214938728,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004917825815617205,
      "loss": 5.1989,
      "mean_token_accuracy": 0.185593119263649,
      "num_tokens": 20832811.0,
      "step": 12005
    },
    {
      "entropy": 5.6006190299987795,
      "epoch": 0.9344485508655903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917751580432229,
      "loss": 5.2671,
      "mean_token_accuracy": 0.1800979867577553,
      "num_tokens": 20841824.0,
      "step": 12010
    },
    {
      "entropy": 5.582976198196411,
      "epoch": 0.9348375802373079,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004917677312355072,
      "loss": 5.2064,
      "mean_token_accuracy": 0.18858443647623063,
      "num_tokens": 20850220.0,
      "step": 12015
    },
    {
      "entropy": 5.487300205230713,
      "epoch": 0.9352266096090255,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917603011386863,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19209174066781998,
      "num_tokens": 20859193.0,
      "step": 12020
    },
    {
      "entropy": 5.491845607757568,
      "epoch": 0.9356156389807431,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917528677528727,
      "loss": 5.1862,
      "mean_token_accuracy": 0.17971764355897904,
      "num_tokens": 20867610.0,
      "step": 12025
    },
    {
      "entropy": 5.6097249507904055,
      "epoch": 0.9360046683524607,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917454310781795,
      "loss": 5.2844,
      "mean_token_accuracy": 0.17608041912317277,
      "num_tokens": 20876991.0,
      "step": 12030
    },
    {
      "entropy": 5.545649671554566,
      "epoch": 0.9363936977241781,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004917379911147193,
      "loss": 5.1416,
      "mean_token_accuracy": 0.18648932576179506,
      "num_tokens": 20885794.0,
      "step": 12035
    },
    {
      "entropy": 5.606280183792114,
      "epoch": 0.9367827270958957,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004917305478626049,
      "loss": 5.3358,
      "mean_token_accuracy": 0.18447591066360475,
      "num_tokens": 20893487.0,
      "step": 12040
    },
    {
      "entropy": 5.595366144180298,
      "epoch": 0.9371717564676133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917231013219495,
      "loss": 5.2703,
      "mean_token_accuracy": 0.17662569433450698,
      "num_tokens": 20902587.0,
      "step": 12045
    },
    {
      "entropy": 5.612043333053589,
      "epoch": 0.9375607858393309,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491715651492866,
      "loss": 5.1933,
      "mean_token_accuracy": 0.19155954867601394,
      "num_tokens": 20911062.0,
      "step": 12050
    },
    {
      "entropy": 5.591382217407227,
      "epoch": 0.9379498152110485,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917081983754675,
      "loss": 5.2279,
      "mean_token_accuracy": 0.18330192863941192,
      "num_tokens": 20919369.0,
      "step": 12055
    },
    {
      "entropy": 5.586066484451294,
      "epoch": 0.938338844582766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004917007419698669,
      "loss": 5.2955,
      "mean_token_accuracy": 0.18256671726703644,
      "num_tokens": 20927514.0,
      "step": 12060
    },
    {
      "entropy": 5.695451641082764,
      "epoch": 0.9387278739544835,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004916932822761776,
      "loss": 5.3385,
      "mean_token_accuracy": 0.1756583884358406,
      "num_tokens": 20935750.0,
      "step": 12065
    },
    {
      "entropy": 5.526162052154541,
      "epoch": 0.9391169033262011,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916858192945126,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18862030655145645,
      "num_tokens": 20944089.0,
      "step": 12070
    },
    {
      "entropy": 5.537595987319946,
      "epoch": 0.9395059326979187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916783530249852,
      "loss": 5.3022,
      "mean_token_accuracy": 0.17719412297010423,
      "num_tokens": 20952704.0,
      "step": 12075
    },
    {
      "entropy": 5.562045240402222,
      "epoch": 0.9398949620696363,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004916708834677086,
      "loss": 5.2016,
      "mean_token_accuracy": 0.17946707457304,
      "num_tokens": 20960741.0,
      "step": 12080
    },
    {
      "entropy": 5.578358125686646,
      "epoch": 0.9402839914413538,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916634106227962,
      "loss": 5.1963,
      "mean_token_accuracy": 0.18899319767951966,
      "num_tokens": 20968778.0,
      "step": 12085
    },
    {
      "entropy": 5.5808814525604244,
      "epoch": 0.9406730208130714,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916559344903614,
      "loss": 5.348,
      "mean_token_accuracy": 0.17608641386032103,
      "num_tokens": 20977993.0,
      "step": 12090
    },
    {
      "entropy": 5.5935197353363035,
      "epoch": 0.941062050184789,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916484550705176,
      "loss": 5.1836,
      "mean_token_accuracy": 0.18955615311861038,
      "num_tokens": 20986974.0,
      "step": 12095
    },
    {
      "entropy": 5.617712116241455,
      "epoch": 0.9414510795565065,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004916409723633785,
      "loss": 5.3067,
      "mean_token_accuracy": 0.1803364261984825,
      "num_tokens": 20995224.0,
      "step": 12100
    },
    {
      "entropy": 5.568239068984985,
      "epoch": 0.9418401089282241,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004916334863690573,
      "loss": 5.3171,
      "mean_token_accuracy": 0.17628771662712098,
      "num_tokens": 21003270.0,
      "step": 12105
    },
    {
      "entropy": 5.570231676101685,
      "epoch": 0.9422291382999416,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004916259970876678,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1799635410308838,
      "num_tokens": 21012665.0,
      "step": 12110
    },
    {
      "entropy": 5.588885307312012,
      "epoch": 0.9426181676716592,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004916185045193237,
      "loss": 5.2998,
      "mean_token_accuracy": 0.1827046051621437,
      "num_tokens": 21021043.0,
      "step": 12115
    },
    {
      "entropy": 5.577015590667725,
      "epoch": 0.9430071970433768,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004916110086641384,
      "loss": 5.2234,
      "mean_token_accuracy": 0.18810662627220154,
      "num_tokens": 21029525.0,
      "step": 12120
    },
    {
      "entropy": 5.548351097106933,
      "epoch": 0.9433962264150944,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004916035095222259,
      "loss": 5.2057,
      "mean_token_accuracy": 0.1850769951939583,
      "num_tokens": 21037687.0,
      "step": 12125
    },
    {
      "entropy": 5.599748611450195,
      "epoch": 0.9437852557868119,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915960070936999,
      "loss": 5.2907,
      "mean_token_accuracy": 0.17675571739673615,
      "num_tokens": 21046095.0,
      "step": 12130
    },
    {
      "entropy": 5.588860702514649,
      "epoch": 0.9441742851585294,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915885013786742,
      "loss": 5.2459,
      "mean_token_accuracy": 0.18637708574533463,
      "num_tokens": 21054565.0,
      "step": 12135
    },
    {
      "entropy": 5.519547414779663,
      "epoch": 0.944563314530247,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915809923772628,
      "loss": 5.1559,
      "mean_token_accuracy": 0.19223669916391373,
      "num_tokens": 21062942.0,
      "step": 12140
    },
    {
      "entropy": 5.577270746231079,
      "epoch": 0.9449523439019646,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004915734800895796,
      "loss": 5.2104,
      "mean_token_accuracy": 0.18142497390508652,
      "num_tokens": 21070642.0,
      "step": 12145
    },
    {
      "entropy": 5.587794208526612,
      "epoch": 0.9453413732736822,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004915659645157383,
      "loss": 5.1878,
      "mean_token_accuracy": 0.18561283349990845,
      "num_tokens": 21079007.0,
      "step": 12150
    },
    {
      "entropy": 5.585548353195191,
      "epoch": 0.9457304026453998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915584456558535,
      "loss": 5.2668,
      "mean_token_accuracy": 0.17792163491249086,
      "num_tokens": 21087672.0,
      "step": 12155
    },
    {
      "entropy": 5.622518682479859,
      "epoch": 0.9461194320171173,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915509235100388,
      "loss": 5.199,
      "mean_token_accuracy": 0.18833868950605392,
      "num_tokens": 21095954.0,
      "step": 12160
    },
    {
      "entropy": 5.609609174728393,
      "epoch": 0.9465084613888348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915433980784086,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1813298523426056,
      "num_tokens": 21104439.0,
      "step": 12165
    },
    {
      "entropy": 5.546779680252075,
      "epoch": 0.9468974907605524,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915358693610769,
      "loss": 5.2176,
      "mean_token_accuracy": 0.17905055731534958,
      "num_tokens": 21112933.0,
      "step": 12170
    },
    {
      "entropy": 5.5651905059814455,
      "epoch": 0.94728652013227,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915283373581581,
      "loss": 5.2174,
      "mean_token_accuracy": 0.19011888206005095,
      "num_tokens": 21122085.0,
      "step": 12175
    },
    {
      "entropy": 5.60441951751709,
      "epoch": 0.9476755495039876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915208020697664,
      "loss": 5.2661,
      "mean_token_accuracy": 0.18176874071359633,
      "num_tokens": 21130994.0,
      "step": 12180
    },
    {
      "entropy": 5.5884778022766115,
      "epoch": 0.9480645788757052,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915132634960162,
      "loss": 5.1947,
      "mean_token_accuracy": 0.18534459322690963,
      "num_tokens": 21139744.0,
      "step": 12185
    },
    {
      "entropy": 5.6022614479064945,
      "epoch": 0.9484536082474226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915057216370218,
      "loss": 5.2009,
      "mean_token_accuracy": 0.18926140367984773,
      "num_tokens": 21148333.0,
      "step": 12190
    },
    {
      "entropy": 5.538546419143676,
      "epoch": 0.9488426376191402,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004914981764928977,
      "loss": 5.1466,
      "mean_token_accuracy": 0.19535300731658936,
      "num_tokens": 21157450.0,
      "step": 12195
    },
    {
      "entropy": 5.598568248748779,
      "epoch": 0.9492316669908578,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004914906280637584,
      "loss": 5.31,
      "mean_token_accuracy": 0.1787262350320816,
      "num_tokens": 21166410.0,
      "step": 12200
    },
    {
      "entropy": 5.587969589233398,
      "epoch": 0.9496206963625754,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914830763497183,
      "loss": 5.2381,
      "mean_token_accuracy": 0.18742359131574632,
      "num_tokens": 21175104.0,
      "step": 12205
    },
    {
      "entropy": 5.578275918960571,
      "epoch": 0.950009725734293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004914755213508922,
      "loss": 5.2013,
      "mean_token_accuracy": 0.18188783377408982,
      "num_tokens": 21182597.0,
      "step": 12210
    },
    {
      "entropy": 5.534088277816773,
      "epoch": 0.9503987551060105,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004914679630673945,
      "loss": 5.2922,
      "mean_token_accuracy": 0.18467426747083665,
      "num_tokens": 21192399.0,
      "step": 12215
    },
    {
      "entropy": 5.566508102416992,
      "epoch": 0.950787784477728,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004914604014993401,
      "loss": 5.1618,
      "mean_token_accuracy": 0.18967350274324418,
      "num_tokens": 21201275.0,
      "step": 12220
    },
    {
      "entropy": 5.585641384124756,
      "epoch": 0.9511768138494456,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004914528366468436,
      "loss": 5.2641,
      "mean_token_accuracy": 0.1857493259012699,
      "num_tokens": 21210491.0,
      "step": 12225
    },
    {
      "entropy": 5.556947231292725,
      "epoch": 0.9515658432211632,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004914452685100199,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18351813703775405,
      "num_tokens": 21218899.0,
      "step": 12230
    },
    {
      "entropy": 5.49579119682312,
      "epoch": 0.9519548725928808,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004914376970889836,
      "loss": 5.1757,
      "mean_token_accuracy": 0.178849957883358,
      "num_tokens": 21227640.0,
      "step": 12235
    },
    {
      "entropy": 5.533745861053466,
      "epoch": 0.9523439019645983,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914301223838497,
      "loss": 5.1392,
      "mean_token_accuracy": 0.18620429337024688,
      "num_tokens": 21236422.0,
      "step": 12240
    },
    {
      "entropy": 5.592987775802612,
      "epoch": 0.9527329313363159,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004914225443947334,
      "loss": 5.3265,
      "mean_token_accuracy": 0.18247947841882706,
      "num_tokens": 21245820.0,
      "step": 12245
    },
    {
      "entropy": 5.644172191619873,
      "epoch": 0.9531219607080335,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914149631217494,
      "loss": 5.2446,
      "mean_token_accuracy": 0.18194653689861298,
      "num_tokens": 21255570.0,
      "step": 12250
    },
    {
      "entropy": 5.622006416320801,
      "epoch": 0.953510990079751,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004914073785650127,
      "loss": 5.3452,
      "mean_token_accuracy": 0.17937376499176025,
      "num_tokens": 21263712.0,
      "step": 12255
    },
    {
      "entropy": 5.611487007141113,
      "epoch": 0.9539000194514686,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913997907246385,
      "loss": 5.2411,
      "mean_token_accuracy": 0.1884733110666275,
      "num_tokens": 21272309.0,
      "step": 12260
    },
    {
      "entropy": 5.488160943984985,
      "epoch": 0.9542890488231861,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000491392199600742,
      "loss": 5.1747,
      "mean_token_accuracy": 0.178478467464447,
      "num_tokens": 21281711.0,
      "step": 12265
    },
    {
      "entropy": 5.535889625549316,
      "epoch": 0.9546780781949037,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004913846051934382,
      "loss": 5.145,
      "mean_token_accuracy": 0.191595858335495,
      "num_tokens": 21289524.0,
      "step": 12270
    },
    {
      "entropy": 5.53168215751648,
      "epoch": 0.9550671075666213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004913770075028425,
      "loss": 5.0872,
      "mean_token_accuracy": 0.19341966658830642,
      "num_tokens": 21298225.0,
      "step": 12275
    },
    {
      "entropy": 5.5428338050842285,
      "epoch": 0.9554561369383389,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004913694065290701,
      "loss": 5.0974,
      "mean_token_accuracy": 0.1931614547967911,
      "num_tokens": 21306819.0,
      "step": 12280
    },
    {
      "entropy": 5.519109725952148,
      "epoch": 0.9558451663100564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004913618022722363,
      "loss": 5.2648,
      "mean_token_accuracy": 0.1792328655719757,
      "num_tokens": 21315517.0,
      "step": 12285
    },
    {
      "entropy": 5.534438419342041,
      "epoch": 0.9562341956817739,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913541947324566,
      "loss": 5.1673,
      "mean_token_accuracy": 0.18894987553358078,
      "num_tokens": 21324144.0,
      "step": 12290
    },
    {
      "entropy": 5.5736658573150635,
      "epoch": 0.9566232250534915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004913465839098463,
      "loss": 5.2244,
      "mean_token_accuracy": 0.18304170668125153,
      "num_tokens": 21333329.0,
      "step": 12295
    },
    {
      "entropy": 5.549862909317016,
      "epoch": 0.9570122544252091,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004913389698045209,
      "loss": 5.1875,
      "mean_token_accuracy": 0.17800935059785844,
      "num_tokens": 21341955.0,
      "step": 12300
    },
    {
      "entropy": 5.547424268722534,
      "epoch": 0.9574012837969267,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491331352416596,
      "loss": 5.1664,
      "mean_token_accuracy": 0.18679251968860627,
      "num_tokens": 21350593.0,
      "step": 12305
    },
    {
      "entropy": 5.582381105422973,
      "epoch": 0.9577903131686443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913237317461872,
      "loss": 5.2462,
      "mean_token_accuracy": 0.18163011819124222,
      "num_tokens": 21359007.0,
      "step": 12310
    },
    {
      "entropy": 5.581271028518676,
      "epoch": 0.9581793425403617,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049131610779341,
      "loss": 5.2869,
      "mean_token_accuracy": 0.17535074651241303,
      "num_tokens": 21367645.0,
      "step": 12315
    },
    {
      "entropy": 5.492307233810425,
      "epoch": 0.9585683719120793,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004913084805583803,
      "loss": 5.08,
      "mean_token_accuracy": 0.18706974238157273,
      "num_tokens": 21376040.0,
      "step": 12320
    },
    {
      "entropy": 5.584387874603271,
      "epoch": 0.9589574012837969,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004913008500412136,
      "loss": 5.2103,
      "mean_token_accuracy": 0.18781665712594986,
      "num_tokens": 21385305.0,
      "step": 12325
    },
    {
      "entropy": 5.534493637084961,
      "epoch": 0.9593464306555145,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912932162420259,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19863148480653764,
      "num_tokens": 21393705.0,
      "step": 12330
    },
    {
      "entropy": 5.5688214778900145,
      "epoch": 0.9597354600272321,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004912855791609327,
      "loss": 5.2366,
      "mean_token_accuracy": 0.1812411978840828,
      "num_tokens": 21401830.0,
      "step": 12335
    },
    {
      "entropy": 5.563437557220459,
      "epoch": 0.9601244893989496,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912779387980502,
      "loss": 5.2679,
      "mean_token_accuracy": 0.18128424286842346,
      "num_tokens": 21410398.0,
      "step": 12340
    },
    {
      "entropy": 5.642089700698852,
      "epoch": 0.9605135187706672,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912702951534942,
      "loss": 5.2415,
      "mean_token_accuracy": 0.18362317085266114,
      "num_tokens": 21419082.0,
      "step": 12345
    },
    {
      "entropy": 5.633106327056884,
      "epoch": 0.9609025481423847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912626482273809,
      "loss": 5.2254,
      "mean_token_accuracy": 0.1831813022494316,
      "num_tokens": 21427992.0,
      "step": 12350
    },
    {
      "entropy": 5.551148223876953,
      "epoch": 0.9612915775141023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912549980198258,
      "loss": 5.2001,
      "mean_token_accuracy": 0.1776973858475685,
      "num_tokens": 21436933.0,
      "step": 12355
    },
    {
      "entropy": 5.539190769195557,
      "epoch": 0.9616806068858199,
      "grad_norm": 1.0,
      "learning_rate": 0.0004912473445309455,
      "loss": 5.2212,
      "mean_token_accuracy": 0.18540673106908798,
      "num_tokens": 21445640.0,
      "step": 12360
    },
    {
      "entropy": 5.638525152206421,
      "epoch": 0.9620696362575375,
      "grad_norm": 1.0,
      "learning_rate": 0.000491239687760856,
      "loss": 5.3556,
      "mean_token_accuracy": 0.17749987095594405,
      "num_tokens": 21454511.0,
      "step": 12365
    },
    {
      "entropy": 5.614642333984375,
      "epoch": 0.962458665629255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912320277096732,
      "loss": 5.3604,
      "mean_token_accuracy": 0.1762730970978737,
      "num_tokens": 21463851.0,
      "step": 12370
    },
    {
      "entropy": 5.5660522937774655,
      "epoch": 0.9628476950009726,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004912243643775137,
      "loss": 5.155,
      "mean_token_accuracy": 0.18747493773698806,
      "num_tokens": 21471620.0,
      "step": 12375
    },
    {
      "entropy": 5.52073826789856,
      "epoch": 0.9632367243726901,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004912166977644936,
      "loss": 5.1773,
      "mean_token_accuracy": 0.18296215534210206,
      "num_tokens": 21479944.0,
      "step": 12380
    },
    {
      "entropy": 5.642212581634522,
      "epoch": 0.9636257537444077,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912090278707293,
      "loss": 5.3355,
      "mean_token_accuracy": 0.17370143234729768,
      "num_tokens": 21489175.0,
      "step": 12385
    },
    {
      "entropy": 5.616637945175171,
      "epoch": 0.9640147831161253,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491201354696337,
      "loss": 5.2516,
      "mean_token_accuracy": 0.1843176707625389,
      "num_tokens": 21498533.0,
      "step": 12390
    },
    {
      "entropy": 5.687803936004639,
      "epoch": 0.9644038124878428,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911936782414334,
      "loss": 5.315,
      "mean_token_accuracy": 0.17314152270555497,
      "num_tokens": 21506771.0,
      "step": 12395
    },
    {
      "entropy": 5.475944852828979,
      "epoch": 0.9647928418595604,
      "grad_norm": 1.0,
      "learning_rate": 0.0004911859985061348,
      "loss": 5.1252,
      "mean_token_accuracy": 0.1967193990945816,
      "num_tokens": 21515359.0,
      "step": 12400
    },
    {
      "entropy": 5.522892618179322,
      "epoch": 0.965181871231278,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004911783154905577,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1812639445066452,
      "num_tokens": 21524313.0,
      "step": 12405
    },
    {
      "entropy": 5.522868347167969,
      "epoch": 0.9655709006029956,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911706291948188,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19219857901334764,
      "num_tokens": 21533211.0,
      "step": 12410
    },
    {
      "entropy": 5.596713399887085,
      "epoch": 0.9659599299747131,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004911629396190348,
      "loss": 5.1528,
      "mean_token_accuracy": 0.18876586705446244,
      "num_tokens": 21541575.0,
      "step": 12415
    },
    {
      "entropy": 5.6097218036651615,
      "epoch": 0.9663489593464306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004911552467633221,
      "loss": 5.2612,
      "mean_token_accuracy": 0.18455181121826172,
      "num_tokens": 21549569.0,
      "step": 12420
    },
    {
      "entropy": 5.5362770557403564,
      "epoch": 0.9667379887181482,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004911475506277976,
      "loss": 5.2052,
      "mean_token_accuracy": 0.180090469121933,
      "num_tokens": 21557277.0,
      "step": 12425
    },
    {
      "entropy": 5.553881359100342,
      "epoch": 0.9671270180898658,
      "grad_norm": 1.046875,
      "learning_rate": 0.000491139851212578,
      "loss": 5.2171,
      "mean_token_accuracy": 0.18169561475515367,
      "num_tokens": 21566590.0,
      "step": 12430
    },
    {
      "entropy": 5.608275032043457,
      "epoch": 0.9675160474615834,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911321485177802,
      "loss": 5.2809,
      "mean_token_accuracy": 0.17754004001617432,
      "num_tokens": 21575140.0,
      "step": 12435
    },
    {
      "entropy": 5.5408426284790036,
      "epoch": 0.967905076833301,
      "grad_norm": 1.0,
      "learning_rate": 0.0004911244425435212,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1739608407020569,
      "num_tokens": 21584647.0,
      "step": 12440
    },
    {
      "entropy": 5.594983911514282,
      "epoch": 0.9682941062050184,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004911167332899176,
      "loss": 5.2039,
      "mean_token_accuracy": 0.18198700547218322,
      "num_tokens": 21593609.0,
      "step": 12445
    },
    {
      "entropy": 5.649124383926392,
      "epoch": 0.968683135576736,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004911090207570867,
      "loss": 5.3282,
      "mean_token_accuracy": 0.18088028877973555,
      "num_tokens": 21602584.0,
      "step": 12450
    },
    {
      "entropy": 5.518971157073975,
      "epoch": 0.9690721649484536,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004911013049451453,
      "loss": 5.2816,
      "mean_token_accuracy": 0.18631130307912827,
      "num_tokens": 21611109.0,
      "step": 12455
    },
    {
      "entropy": 5.6388171195983885,
      "epoch": 0.9694611943201712,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910935858542106,
      "loss": 5.2748,
      "mean_token_accuracy": 0.18172716796398164,
      "num_tokens": 21620173.0,
      "step": 12460
    },
    {
      "entropy": 5.530540513992309,
      "epoch": 0.9698502236918888,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004910858634843997,
      "loss": 5.1228,
      "mean_token_accuracy": 0.1912248730659485,
      "num_tokens": 21628318.0,
      "step": 12465
    },
    {
      "entropy": 5.615167903900146,
      "epoch": 0.9702392530636063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910781378358297,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18239201456308365,
      "num_tokens": 21637102.0,
      "step": 12470
    },
    {
      "entropy": 5.58656063079834,
      "epoch": 0.9706282824353238,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004910704089086178,
      "loss": 5.1829,
      "mean_token_accuracy": 0.1865543931722641,
      "num_tokens": 21645055.0,
      "step": 12475
    },
    {
      "entropy": 5.58342022895813,
      "epoch": 0.9710173118070414,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910626767028815,
      "loss": 5.2381,
      "mean_token_accuracy": 0.1781092554330826,
      "num_tokens": 21654106.0,
      "step": 12480
    },
    {
      "entropy": 5.568834114074707,
      "epoch": 0.971406341178759,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004910549412187379,
      "loss": 5.2366,
      "mean_token_accuracy": 0.18529815524816512,
      "num_tokens": 21663081.0,
      "step": 12485
    },
    {
      "entropy": 5.5896697521209715,
      "epoch": 0.9717953705504766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004910472024563045,
      "loss": 5.2048,
      "mean_token_accuracy": 0.18465072959661483,
      "num_tokens": 21671033.0,
      "step": 12490
    },
    {
      "entropy": 5.5985438346862795,
      "epoch": 0.9721843999221941,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004910394604156987,
      "loss": 5.2134,
      "mean_token_accuracy": 0.18573009669780732,
      "num_tokens": 21679167.0,
      "step": 12495
    },
    {
      "entropy": 5.588705062866211,
      "epoch": 0.9725734292939117,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910317150970379,
      "loss": 5.3238,
      "mean_token_accuracy": 0.1814632698893547,
      "num_tokens": 21688365.0,
      "step": 12500
    },
    {
      "entropy": 5.539926433563233,
      "epoch": 0.9729624586656292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910239665004397,
      "loss": 5.2235,
      "mean_token_accuracy": 0.18637812435626983,
      "num_tokens": 21697003.0,
      "step": 12505
    },
    {
      "entropy": 5.597105979919434,
      "epoch": 0.9733514880373468,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004910162146260216,
      "loss": 5.2173,
      "mean_token_accuracy": 0.18418617844581603,
      "num_tokens": 21705947.0,
      "step": 12510
    },
    {
      "entropy": 5.612230205535889,
      "epoch": 0.9737405174090644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004910084594739013,
      "loss": 5.2608,
      "mean_token_accuracy": 0.17966595143079758,
      "num_tokens": 21714153.0,
      "step": 12515
    },
    {
      "entropy": 5.604459524154663,
      "epoch": 0.9741295467807819,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004910007010441964,
      "loss": 5.3203,
      "mean_token_accuracy": 0.1762784644961357,
      "num_tokens": 21723480.0,
      "step": 12520
    },
    {
      "entropy": 5.646980857849121,
      "epoch": 0.9745185761524995,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004909929393370248,
      "loss": 5.1961,
      "mean_token_accuracy": 0.18544194400310515,
      "num_tokens": 21731297.0,
      "step": 12525
    },
    {
      "entropy": 5.551740550994873,
      "epoch": 0.9749076055242171,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909851743525041,
      "loss": 5.1858,
      "mean_token_accuracy": 0.18293648213148117,
      "num_tokens": 21740975.0,
      "step": 12530
    },
    {
      "entropy": 5.496854162216186,
      "epoch": 0.9752966348959347,
      "grad_norm": 1.03125,
      "learning_rate": 0.000490977406090752,
      "loss": 5.1778,
      "mean_token_accuracy": 0.18378167897462844,
      "num_tokens": 21749642.0,
      "step": 12535
    },
    {
      "entropy": 5.548439884185791,
      "epoch": 0.9756856642676522,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004909696345518867,
      "loss": 5.2225,
      "mean_token_accuracy": 0.18318607807159423,
      "num_tokens": 21757962.0,
      "step": 12540
    },
    {
      "entropy": 5.598895406723022,
      "epoch": 0.9760746936393697,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909618597360258,
      "loss": 5.2093,
      "mean_token_accuracy": 0.18401122093200684,
      "num_tokens": 21766053.0,
      "step": 12545
    },
    {
      "entropy": 5.522032165527344,
      "epoch": 0.9764637230110873,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004909540816432876,
      "loss": 5.1679,
      "mean_token_accuracy": 0.18140005469322204,
      "num_tokens": 21774280.0,
      "step": 12550
    },
    {
      "entropy": 5.52596960067749,
      "epoch": 0.9768527523828049,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004909463002737897,
      "loss": 5.1999,
      "mean_token_accuracy": 0.1816786602139473,
      "num_tokens": 21782879.0,
      "step": 12555
    },
    {
      "entropy": 5.5955650329589846,
      "epoch": 0.9772417817545225,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004909385156276506,
      "loss": 5.2781,
      "mean_token_accuracy": 0.18505041748285295,
      "num_tokens": 21791294.0,
      "step": 12560
    },
    {
      "entropy": 5.580241298675537,
      "epoch": 0.9776308111262401,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004909307277049881,
      "loss": 5.1347,
      "mean_token_accuracy": 0.18864893317222595,
      "num_tokens": 21799304.0,
      "step": 12565
    },
    {
      "entropy": 5.588964462280273,
      "epoch": 0.9780198404979576,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004909229365059205,
      "loss": 5.3029,
      "mean_token_accuracy": 0.17473739236593247,
      "num_tokens": 21809024.0,
      "step": 12570
    },
    {
      "entropy": 5.566770839691162,
      "epoch": 0.9784088698696751,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004909151420305661,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19593019485473634,
      "num_tokens": 21817412.0,
      "step": 12575
    },
    {
      "entropy": 5.516671133041382,
      "epoch": 0.9787978992413927,
      "grad_norm": 1.09375,
      "learning_rate": 0.000490907344279043,
      "loss": 5.1664,
      "mean_token_accuracy": 0.18375322371721267,
      "num_tokens": 21825261.0,
      "step": 12580
    },
    {
      "entropy": 5.6050732135772705,
      "epoch": 0.9791869286131103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908995432514698,
      "loss": 5.2775,
      "mean_token_accuracy": 0.18813448548316955,
      "num_tokens": 21833318.0,
      "step": 12585
    },
    {
      "entropy": 5.585233592987061,
      "epoch": 0.9795759579848279,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908917389479645,
      "loss": 5.239,
      "mean_token_accuracy": 0.18393113762140273,
      "num_tokens": 21841377.0,
      "step": 12590
    },
    {
      "entropy": 5.5364807605743405,
      "epoch": 0.9799649873565455,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004908839313686456,
      "loss": 5.185,
      "mean_token_accuracy": 0.1848248153924942,
      "num_tokens": 21850183.0,
      "step": 12595
    },
    {
      "entropy": 5.599163150787353,
      "epoch": 0.980354016728263,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004908761205136319,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19129650443792343,
      "num_tokens": 21858108.0,
      "step": 12600
    },
    {
      "entropy": 5.546742153167725,
      "epoch": 0.9807430460999805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908683063830414,
      "loss": 5.1664,
      "mean_token_accuracy": 0.18518624752759932,
      "num_tokens": 21866330.0,
      "step": 12605
    },
    {
      "entropy": 5.580180692672729,
      "epoch": 0.9811320754716981,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004908604889769932,
      "loss": 5.2354,
      "mean_token_accuracy": 0.18176840841770173,
      "num_tokens": 21875254.0,
      "step": 12610
    },
    {
      "entropy": 5.599694490432739,
      "epoch": 0.9815211048434157,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908526682956054,
      "loss": 5.2965,
      "mean_token_accuracy": 0.18086742162704467,
      "num_tokens": 21884009.0,
      "step": 12615
    },
    {
      "entropy": 5.598038005828857,
      "epoch": 0.9819101342151333,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004908448443389972,
      "loss": 5.2147,
      "mean_token_accuracy": 0.17888280302286147,
      "num_tokens": 21892428.0,
      "step": 12620
    },
    {
      "entropy": 5.596229982376099,
      "epoch": 0.9822991635868508,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004908370171072869,
      "loss": 5.251,
      "mean_token_accuracy": 0.19143105298280716,
      "num_tokens": 21901140.0,
      "step": 12625
    },
    {
      "entropy": 5.526429319381714,
      "epoch": 0.9826881929585684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908291866005933,
      "loss": 5.1563,
      "mean_token_accuracy": 0.19609212428331374,
      "num_tokens": 21909072.0,
      "step": 12630
    },
    {
      "entropy": 5.543815803527832,
      "epoch": 0.9830772223302859,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908213528190355,
      "loss": 5.2264,
      "mean_token_accuracy": 0.1802401065826416,
      "num_tokens": 21917873.0,
      "step": 12635
    },
    {
      "entropy": 5.6887753963470455,
      "epoch": 0.9834662517020035,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004908135157627321,
      "loss": 5.3186,
      "mean_token_accuracy": 0.17546849995851516,
      "num_tokens": 21927302.0,
      "step": 12640
    },
    {
      "entropy": 5.598447847366333,
      "epoch": 0.9838552810737211,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908056754318022,
      "loss": 5.2572,
      "mean_token_accuracy": 0.1846102222800255,
      "num_tokens": 21936440.0,
      "step": 12645
    },
    {
      "entropy": 5.551378440856934,
      "epoch": 0.9842443104454386,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004907978318263646,
      "loss": 5.1262,
      "mean_token_accuracy": 0.18936778604984283,
      "num_tokens": 21944655.0,
      "step": 12650
    },
    {
      "entropy": 5.694837379455566,
      "epoch": 0.9846333398171562,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004907899849465383,
      "loss": 5.3495,
      "mean_token_accuracy": 0.17975814491510392,
      "num_tokens": 21953202.0,
      "step": 12655
    },
    {
      "entropy": 5.615654182434082,
      "epoch": 0.9850223691888738,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004907821347924424,
      "loss": 5.2056,
      "mean_token_accuracy": 0.18635283261537552,
      "num_tokens": 21961889.0,
      "step": 12660
    },
    {
      "entropy": 5.5188713550567625,
      "epoch": 0.9854113985605913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907742813641963,
      "loss": 5.1053,
      "mean_token_accuracy": 0.1899982675909996,
      "num_tokens": 21970864.0,
      "step": 12665
    },
    {
      "entropy": 5.5535773754119875,
      "epoch": 0.9858004279323089,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004907664246619187,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18352774828672408,
      "num_tokens": 21979271.0,
      "step": 12670
    },
    {
      "entropy": 5.588700294494629,
      "epoch": 0.9861894573040264,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004907585646857293,
      "loss": 5.0586,
      "mean_token_accuracy": 0.19708133935928346,
      "num_tokens": 21987321.0,
      "step": 12675
    },
    {
      "entropy": 5.578938150405884,
      "epoch": 0.986578486675744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004907507014357467,
      "loss": 5.3289,
      "mean_token_accuracy": 0.18304667323827745,
      "num_tokens": 21996094.0,
      "step": 12680
    },
    {
      "entropy": 5.478115272521973,
      "epoch": 0.9869675160474616,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004907428349120909,
      "loss": 5.1776,
      "mean_token_accuracy": 0.18315355032682418,
      "num_tokens": 22004095.0,
      "step": 12685
    },
    {
      "entropy": 5.633507013320923,
      "epoch": 0.9873565454191792,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004907349651148809,
      "loss": 5.2887,
      "mean_token_accuracy": 0.18042635023593903,
      "num_tokens": 22013371.0,
      "step": 12690
    },
    {
      "entropy": 5.615620231628418,
      "epoch": 0.9877455747908968,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004907270920442361,
      "loss": 5.1597,
      "mean_token_accuracy": 0.18618180751800537,
      "num_tokens": 22022313.0,
      "step": 12695
    },
    {
      "entropy": 5.606718397140503,
      "epoch": 0.9881346041626142,
      "grad_norm": 1.0,
      "learning_rate": 0.0004907192157002762,
      "loss": 5.3242,
      "mean_token_accuracy": 0.1854074165225029,
      "num_tokens": 22031690.0,
      "step": 12700
    },
    {
      "entropy": 5.697993612289428,
      "epoch": 0.9885236335343318,
      "grad_norm": 1.125,
      "learning_rate": 0.0004907113360831204,
      "loss": 5.3687,
      "mean_token_accuracy": 0.17096651196479798,
      "num_tokens": 22040957.0,
      "step": 12705
    },
    {
      "entropy": 5.628562545776367,
      "epoch": 0.9889126629060494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004907034531928886,
      "loss": 5.2044,
      "mean_token_accuracy": 0.18460477739572526,
      "num_tokens": 22049623.0,
      "step": 12710
    },
    {
      "entropy": 5.564844226837158,
      "epoch": 0.989301692277767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906955670297001,
      "loss": 5.1957,
      "mean_token_accuracy": 0.1872323453426361,
      "num_tokens": 22058205.0,
      "step": 12715
    },
    {
      "entropy": 5.537142944335938,
      "epoch": 0.9896907216494846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004906876775936746,
      "loss": 5.0722,
      "mean_token_accuracy": 0.18995586633682252,
      "num_tokens": 22067375.0,
      "step": 12720
    },
    {
      "entropy": 5.540398025512696,
      "epoch": 0.990079751021202,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004906797848849321,
      "loss": 5.119,
      "mean_token_accuracy": 0.19641171246767045,
      "num_tokens": 22075670.0,
      "step": 12725
    },
    {
      "entropy": 5.4911055088043215,
      "epoch": 0.9904687803929196,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000490671888903592,
      "loss": 5.2117,
      "mean_token_accuracy": 0.1884309858083725,
      "num_tokens": 22084360.0,
      "step": 12730
    },
    {
      "entropy": 5.545493793487549,
      "epoch": 0.9908578097646372,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004906639896497744,
      "loss": 5.2247,
      "mean_token_accuracy": 0.19025198519229888,
      "num_tokens": 22092666.0,
      "step": 12735
    },
    {
      "entropy": 5.667842674255371,
      "epoch": 0.9912468391363548,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004906560871235988,
      "loss": 5.3501,
      "mean_token_accuracy": 0.1804563194513321,
      "num_tokens": 22101311.0,
      "step": 12740
    },
    {
      "entropy": 5.603699779510498,
      "epoch": 0.9916358685080724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004906481813251854,
      "loss": 5.1851,
      "mean_token_accuracy": 0.19003836959600448,
      "num_tokens": 22109640.0,
      "step": 12745
    },
    {
      "entropy": 5.525662231445312,
      "epoch": 0.9920248978797899,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906402722546542,
      "loss": 5.154,
      "mean_token_accuracy": 0.18206890672445297,
      "num_tokens": 22118721.0,
      "step": 12750
    },
    {
      "entropy": 5.50973539352417,
      "epoch": 0.9924139272515075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490632359912125,
      "loss": 5.1813,
      "mean_token_accuracy": 0.18638604879379272,
      "num_tokens": 22126745.0,
      "step": 12755
    },
    {
      "entropy": 5.602630090713501,
      "epoch": 0.992802956623225,
      "grad_norm": 1.078125,
      "learning_rate": 0.000490624444297718,
      "loss": 5.2765,
      "mean_token_accuracy": 0.17993869483470917,
      "num_tokens": 22135010.0,
      "step": 12760
    },
    {
      "entropy": 5.60313777923584,
      "epoch": 0.9931919859949426,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004906165254115533,
      "loss": 5.1973,
      "mean_token_accuracy": 0.19051640331745148,
      "num_tokens": 22144339.0,
      "step": 12765
    },
    {
      "entropy": 5.493089199066162,
      "epoch": 0.9935810153666602,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490608603253751,
      "loss": 5.1557,
      "mean_token_accuracy": 0.19175340384244918,
      "num_tokens": 22153405.0,
      "step": 12770
    },
    {
      "entropy": 5.584712123870849,
      "epoch": 0.9939700447383778,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004906006778244312,
      "loss": 5.2214,
      "mean_token_accuracy": 0.18643719404935838,
      "num_tokens": 22161392.0,
      "step": 12775
    },
    {
      "entropy": 5.588856267929077,
      "epoch": 0.9943590741100953,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905927491237145,
      "loss": 5.1886,
      "mean_token_accuracy": 0.1861453667283058,
      "num_tokens": 22170635.0,
      "step": 12780
    },
    {
      "entropy": 5.5074944496154785,
      "epoch": 0.9947481034818129,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905848171517208,
      "loss": 5.0983,
      "mean_token_accuracy": 0.18969143629074098,
      "num_tokens": 22178629.0,
      "step": 12785
    },
    {
      "entropy": 5.5129903793334964,
      "epoch": 0.9951371328535304,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905768819085706,
      "loss": 5.1479,
      "mean_token_accuracy": 0.1912933737039566,
      "num_tokens": 22187142.0,
      "step": 12790
    },
    {
      "entropy": 5.61646876335144,
      "epoch": 0.995526162225248,
      "grad_norm": 1.0,
      "learning_rate": 0.0004905689433943846,
      "loss": 5.2345,
      "mean_token_accuracy": 0.17937208712100983,
      "num_tokens": 22195249.0,
      "step": 12795
    },
    {
      "entropy": 5.6081459522247314,
      "epoch": 0.9959151915969656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004905610016092828,
      "loss": 5.2673,
      "mean_token_accuracy": 0.18590012639760972,
      "num_tokens": 22203621.0,
      "step": 12800
    },
    {
      "entropy": 5.529088068008423,
      "epoch": 0.9963042209686831,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004905530565533859,
      "loss": 5.1976,
      "mean_token_accuracy": 0.1846420183777809,
      "num_tokens": 22212547.0,
      "step": 12805
    },
    {
      "entropy": 5.6065412044525145,
      "epoch": 0.9966932503404007,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905451082268146,
      "loss": 5.2159,
      "mean_token_accuracy": 0.1758275732398033,
      "num_tokens": 22221642.0,
      "step": 12810
    },
    {
      "entropy": 5.573036956787109,
      "epoch": 0.9970822797121183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905371566296891,
      "loss": 5.2399,
      "mean_token_accuracy": 0.17842131108045578,
      "num_tokens": 22230227.0,
      "step": 12815
    },
    {
      "entropy": 5.546908330917359,
      "epoch": 0.9974713090838359,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004905292017621305,
      "loss": 5.2024,
      "mean_token_accuracy": 0.18897558450698854,
      "num_tokens": 22238973.0,
      "step": 12820
    },
    {
      "entropy": 5.548635864257813,
      "epoch": 0.9978603384555534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004905212436242593,
      "loss": 5.177,
      "mean_token_accuracy": 0.19114883989095688,
      "num_tokens": 22247613.0,
      "step": 12825
    },
    {
      "entropy": 5.503739261627198,
      "epoch": 0.9982493678272709,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004905132822161962,
      "loss": 5.224,
      "mean_token_accuracy": 0.17789334803819656,
      "num_tokens": 22255707.0,
      "step": 12830
    },
    {
      "entropy": 5.523511791229248,
      "epoch": 0.9986383971989885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004905053175380621,
      "loss": 5.1243,
      "mean_token_accuracy": 0.19469624161720275,
      "num_tokens": 22264524.0,
      "step": 12835
    },
    {
      "entropy": 5.621120643615723,
      "epoch": 0.9990274265707061,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004904973495899778,
      "loss": 5.2137,
      "mean_token_accuracy": 0.1824513480067253,
      "num_tokens": 22272973.0,
      "step": 12840
    },
    {
      "entropy": 5.6202905654907225,
      "epoch": 0.9994164559424237,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004904893783720642,
      "loss": 5.2442,
      "mean_token_accuracy": 0.18305327892303466,
      "num_tokens": 22281545.0,
      "step": 12845
    },
    {
      "entropy": 5.596254968643189,
      "epoch": 0.9998054853141413,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004904814038844424,
      "loss": 5.2225,
      "mean_token_accuracy": 0.19208308160305024,
      "num_tokens": 22289835.0,
      "step": 12850
    },
    {
      "entropy": 5.551749335394965,
      "epoch": 1.000155611748687,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904734261272332,
      "loss": 5.1947,
      "mean_token_accuracy": 0.18022184736198849,
      "num_tokens": 22297794.0,
      "step": 12855
    },
    {
      "entropy": 5.527487802505493,
      "epoch": 1.0005446411204046,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004904654451005576,
      "loss": 5.1164,
      "mean_token_accuracy": 0.18346790075302125,
      "num_tokens": 22305637.0,
      "step": 12860
    },
    {
      "entropy": 5.591954135894776,
      "epoch": 1.0009336704921221,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904574608045369,
      "loss": 5.1735,
      "mean_token_accuracy": 0.1810388058423996,
      "num_tokens": 22314796.0,
      "step": 12865
    },
    {
      "entropy": 5.582755994796753,
      "epoch": 1.0013226998638398,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004904494732392923,
      "loss": 5.1203,
      "mean_token_accuracy": 0.1836061120033264,
      "num_tokens": 22324522.0,
      "step": 12870
    },
    {
      "entropy": 5.589377593994141,
      "epoch": 1.0017117292355573,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004904414824049448,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19205431491136551,
      "num_tokens": 22333404.0,
      "step": 12875
    },
    {
      "entropy": 5.551284456253052,
      "epoch": 1.0021007586072748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904334883016156,
      "loss": 5.2173,
      "mean_token_accuracy": 0.18167387694120407,
      "num_tokens": 22343364.0,
      "step": 12880
    },
    {
      "entropy": 5.6346841812133786,
      "epoch": 1.0024897879789925,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004904254909294261,
      "loss": 5.1575,
      "mean_token_accuracy": 0.18674262315034867,
      "num_tokens": 22352854.0,
      "step": 12885
    },
    {
      "entropy": 5.572790622711182,
      "epoch": 1.00287881735071,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904174902884978,
      "loss": 5.0557,
      "mean_token_accuracy": 0.19601039141416549,
      "num_tokens": 22361916.0,
      "step": 12890
    },
    {
      "entropy": 5.525163984298706,
      "epoch": 1.0032678467224276,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004904094863789519,
      "loss": 5.143,
      "mean_token_accuracy": 0.19042318314313889,
      "num_tokens": 22370848.0,
      "step": 12895
    },
    {
      "entropy": 5.503107166290283,
      "epoch": 1.0036568760941451,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049040147920091,
      "loss": 5.0568,
      "mean_token_accuracy": 0.19201369434595109,
      "num_tokens": 22379945.0,
      "step": 12900
    },
    {
      "entropy": 5.569567489624023,
      "epoch": 1.0040459054658626,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004903934687544933,
      "loss": 5.1027,
      "mean_token_accuracy": 0.18881996721029282,
      "num_tokens": 22388795.0,
      "step": 12905
    },
    {
      "entropy": 5.575537300109863,
      "epoch": 1.0044349348375803,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004903854550398237,
      "loss": 5.1447,
      "mean_token_accuracy": 0.18939559608697892,
      "num_tokens": 22396960.0,
      "step": 12910
    },
    {
      "entropy": 5.57019009590149,
      "epoch": 1.0048239642092978,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004903774380570226,
      "loss": 5.0261,
      "mean_token_accuracy": 0.19442723840475082,
      "num_tokens": 22404698.0,
      "step": 12915
    },
    {
      "entropy": 5.544312286376953,
      "epoch": 1.0052129935810155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903694178062117,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18495150059461593,
      "num_tokens": 22413391.0,
      "step": 12920
    },
    {
      "entropy": 5.579765272140503,
      "epoch": 1.005602022952733,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004903613942875126,
      "loss": 5.1851,
      "mean_token_accuracy": 0.18706101626157762,
      "num_tokens": 22421357.0,
      "step": 12925
    },
    {
      "entropy": 5.5321300506591795,
      "epoch": 1.0059910523244504,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004903533675010472,
      "loss": 5.0525,
      "mean_token_accuracy": 0.19058032035827638,
      "num_tokens": 22430286.0,
      "step": 12930
    },
    {
      "entropy": 5.5452333927154545,
      "epoch": 1.006380081696168,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903453374469373,
      "loss": 5.1215,
      "mean_token_accuracy": 0.1931103825569153,
      "num_tokens": 22439128.0,
      "step": 12935
    },
    {
      "entropy": 5.529053974151611,
      "epoch": 1.0067691110678856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004903373041253045,
      "loss": 5.0684,
      "mean_token_accuracy": 0.189258474111557,
      "num_tokens": 22447724.0,
      "step": 12940
    },
    {
      "entropy": 5.502752017974854,
      "epoch": 1.0071581404396033,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004903292675362709,
      "loss": 5.1065,
      "mean_token_accuracy": 0.19186557233333587,
      "num_tokens": 22456792.0,
      "step": 12945
    },
    {
      "entropy": 5.456889009475708,
      "epoch": 1.0075471698113208,
      "grad_norm": 1.125,
      "learning_rate": 0.0004903212276799584,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19512889534235,
      "num_tokens": 22465465.0,
      "step": 12950
    },
    {
      "entropy": 5.52273211479187,
      "epoch": 1.0079361991830382,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004903131845564889,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19039687812328338,
      "num_tokens": 22474498.0,
      "step": 12955
    },
    {
      "entropy": 5.5171901226043705,
      "epoch": 1.008325228554756,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903051381659847,
      "loss": 5.0334,
      "mean_token_accuracy": 0.19545533657073974,
      "num_tokens": 22482926.0,
      "step": 12960
    },
    {
      "entropy": 5.528304719924927,
      "epoch": 1.0087142579264734,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004902970885085677,
      "loss": 5.053,
      "mean_token_accuracy": 0.1960665240883827,
      "num_tokens": 22491984.0,
      "step": 12965
    },
    {
      "entropy": 5.5467771053314205,
      "epoch": 1.009103287298191,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049028903558436,
      "loss": 5.0807,
      "mean_token_accuracy": 0.18490277528762816,
      "num_tokens": 22501411.0,
      "step": 12970
    },
    {
      "entropy": 5.582443284988403,
      "epoch": 1.0094923166699086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004902809793934838,
      "loss": 5.2055,
      "mean_token_accuracy": 0.18428508937358856,
      "num_tokens": 22510289.0,
      "step": 12975
    },
    {
      "entropy": 5.612055158615112,
      "epoch": 1.009881346041626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004902729199360615,
      "loss": 5.2273,
      "mean_token_accuracy": 0.18126769810914994,
      "num_tokens": 22519061.0,
      "step": 12980
    },
    {
      "entropy": 5.618602991104126,
      "epoch": 1.0102703754133437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004902648572122153,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1796962559223175,
      "num_tokens": 22528207.0,
      "step": 12985
    },
    {
      "entropy": 5.620871639251709,
      "epoch": 1.0106594047850612,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004902567912220674,
      "loss": 5.1629,
      "mean_token_accuracy": 0.18543503433465958,
      "num_tokens": 22537055.0,
      "step": 12990
    },
    {
      "entropy": 5.518866300582886,
      "epoch": 1.011048434156779,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004902487219657404,
      "loss": 5.1312,
      "mean_token_accuracy": 0.18649035990238189,
      "num_tokens": 22545239.0,
      "step": 12995
    },
    {
      "entropy": 5.5861328601837155,
      "epoch": 1.0114374635284964,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902406494433566,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1743037223815918,
      "num_tokens": 22554850.0,
      "step": 13000
    },
    {
      "entropy": 5.546302080154419,
      "epoch": 1.0118264929002139,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004902325736550386,
      "loss": 5.1076,
      "mean_token_accuracy": 0.18960790038108827,
      "num_tokens": 22563799.0,
      "step": 13005
    },
    {
      "entropy": 5.600475597381592,
      "epoch": 1.0122155222719316,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004902244946009088,
      "loss": 5.3103,
      "mean_token_accuracy": 0.1846143439412117,
      "num_tokens": 22572699.0,
      "step": 13010
    },
    {
      "entropy": 5.668517780303955,
      "epoch": 1.012604551643649,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902164122810899,
      "loss": 5.2403,
      "mean_token_accuracy": 0.17857159525156022,
      "num_tokens": 22580916.0,
      "step": 13015
    },
    {
      "entropy": 5.584632825851441,
      "epoch": 1.0129935810153667,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004902083266957045,
      "loss": 5.1052,
      "mean_token_accuracy": 0.19509144276380538,
      "num_tokens": 22589525.0,
      "step": 13020
    },
    {
      "entropy": 5.590075826644897,
      "epoch": 1.0133826103870842,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902002378448753,
      "loss": 5.1396,
      "mean_token_accuracy": 0.1907185584306717,
      "num_tokens": 22597929.0,
      "step": 13025
    },
    {
      "entropy": 5.587919998168945,
      "epoch": 1.0137716397588017,
      "grad_norm": 1.03125,
      "learning_rate": 0.000490192145728725,
      "loss": 5.1456,
      "mean_token_accuracy": 0.1881784439086914,
      "num_tokens": 22606500.0,
      "step": 13030
    },
    {
      "entropy": 5.5851402759552,
      "epoch": 1.0141606691305194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901840503473764,
      "loss": 5.1136,
      "mean_token_accuracy": 0.18745359033346176,
      "num_tokens": 22614626.0,
      "step": 13035
    },
    {
      "entropy": 5.571524906158447,
      "epoch": 1.0145496985022369,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901759517009522,
      "loss": 5.1588,
      "mean_token_accuracy": 0.17972389310598375,
      "num_tokens": 22623488.0,
      "step": 13040
    },
    {
      "entropy": 5.516906356811523,
      "epoch": 1.0149387278739546,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004901678497895755,
      "loss": 5.0691,
      "mean_token_accuracy": 0.1938052162528038,
      "num_tokens": 22632413.0,
      "step": 13045
    },
    {
      "entropy": 5.617141342163086,
      "epoch": 1.015327757245672,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004901597446133692,
      "loss": 5.2221,
      "mean_token_accuracy": 0.18159100264310837,
      "num_tokens": 22640982.0,
      "step": 13050
    },
    {
      "entropy": 5.587492895126343,
      "epoch": 1.0157167866173895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901516361724564,
      "loss": 5.1692,
      "mean_token_accuracy": 0.18100242763757707,
      "num_tokens": 22649809.0,
      "step": 13055
    },
    {
      "entropy": 5.589676237106323,
      "epoch": 1.0161058159891072,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004901435244669597,
      "loss": 5.0956,
      "mean_token_accuracy": 0.18943627923727036,
      "num_tokens": 22658292.0,
      "step": 13060
    },
    {
      "entropy": 5.652193212509156,
      "epoch": 1.0164948453608247,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901354094970025,
      "loss": 5.171,
      "mean_token_accuracy": 0.18510718643665314,
      "num_tokens": 22667063.0,
      "step": 13065
    },
    {
      "entropy": 5.533492565155029,
      "epoch": 1.0168838747325424,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004901272912627081,
      "loss": 5.1169,
      "mean_token_accuracy": 0.18823547959327697,
      "num_tokens": 22676213.0,
      "step": 13070
    },
    {
      "entropy": 5.45023627281189,
      "epoch": 1.0172729041042599,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004901191697641992,
      "loss": 5.0313,
      "mean_token_accuracy": 0.191065913438797,
      "num_tokens": 22684948.0,
      "step": 13075
    },
    {
      "entropy": 5.573294734954834,
      "epoch": 1.0176619334759773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901110450015994,
      "loss": 5.1425,
      "mean_token_accuracy": 0.18928076177835465,
      "num_tokens": 22693218.0,
      "step": 13080
    },
    {
      "entropy": 5.509079027175903,
      "epoch": 1.018050962847695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901029169750319,
      "loss": 5.0112,
      "mean_token_accuracy": 0.19439788311719894,
      "num_tokens": 22701572.0,
      "step": 13085
    },
    {
      "entropy": 5.526971292495728,
      "epoch": 1.0184399922194125,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004900947856846201,
      "loss": 5.0836,
      "mean_token_accuracy": 0.19143322259187698,
      "num_tokens": 22709498.0,
      "step": 13090
    },
    {
      "entropy": 5.616780376434326,
      "epoch": 1.0188290215911302,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000490086651130487,
      "loss": 5.1117,
      "mean_token_accuracy": 0.18606079667806624,
      "num_tokens": 22717922.0,
      "step": 13095
    },
    {
      "entropy": 5.582559061050415,
      "epoch": 1.0192180509628477,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004900785133127566,
      "loss": 5.1189,
      "mean_token_accuracy": 0.19332155883312224,
      "num_tokens": 22725933.0,
      "step": 13100
    },
    {
      "entropy": 5.5245260238647464,
      "epoch": 1.0196070803345652,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000490070372231552,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19024661928415298,
      "num_tokens": 22734198.0,
      "step": 13105
    },
    {
      "entropy": 5.54635009765625,
      "epoch": 1.0199961097062828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900622278869968,
      "loss": 5.1392,
      "mean_token_accuracy": 0.19009800702333451,
      "num_tokens": 22742419.0,
      "step": 13110
    },
    {
      "entropy": 5.585974550247192,
      "epoch": 1.0203851390780003,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004900540802792146,
      "loss": 5.1048,
      "mean_token_accuracy": 0.19296212941408158,
      "num_tokens": 22750576.0,
      "step": 13115
    },
    {
      "entropy": 5.600459003448487,
      "epoch": 1.020774168449718,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000490045929408329,
      "loss": 5.26,
      "mean_token_accuracy": 0.18776963502168656,
      "num_tokens": 22759604.0,
      "step": 13120
    },
    {
      "entropy": 5.547974538803101,
      "epoch": 1.0211631978214355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900377752744637,
      "loss": 5.0294,
      "mean_token_accuracy": 0.19169464856386184,
      "num_tokens": 22768305.0,
      "step": 13125
    },
    {
      "entropy": 5.625608921051025,
      "epoch": 1.021552227193153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900296178777426,
      "loss": 5.1625,
      "mean_token_accuracy": 0.17878606170415878,
      "num_tokens": 22777364.0,
      "step": 13130
    },
    {
      "entropy": 5.5349259853363035,
      "epoch": 1.0219412565648707,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000490021457218289,
      "loss": 5.0933,
      "mean_token_accuracy": 0.18982576727867126,
      "num_tokens": 22786192.0,
      "step": 13135
    },
    {
      "entropy": 5.653875207901001,
      "epoch": 1.0223302859365881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900132932962272,
      "loss": 5.1624,
      "mean_token_accuracy": 0.19149174094200133,
      "num_tokens": 22794244.0,
      "step": 13140
    },
    {
      "entropy": 5.555508995056153,
      "epoch": 1.0227193153083058,
      "grad_norm": 1.0,
      "learning_rate": 0.000490005126111681,
      "loss": 5.1229,
      "mean_token_accuracy": 0.19354842752218246,
      "num_tokens": 22802543.0,
      "step": 13145
    },
    {
      "entropy": 5.494833374023438,
      "epoch": 1.0231083446800233,
      "grad_norm": 1.0,
      "learning_rate": 0.000489996955664774,
      "loss": 5.157,
      "mean_token_accuracy": 0.18780585676431655,
      "num_tokens": 22811267.0,
      "step": 13150
    },
    {
      "entropy": 5.5426812171936035,
      "epoch": 1.023497374051741,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004899887819556306,
      "loss": 5.0393,
      "mean_token_accuracy": 0.19610452353954316,
      "num_tokens": 22820196.0,
      "step": 13155
    },
    {
      "entropy": 5.615659570693969,
      "epoch": 1.0238864034234585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004899806049843745,
      "loss": 5.2169,
      "mean_token_accuracy": 0.1824982598423958,
      "num_tokens": 22828498.0,
      "step": 13160
    },
    {
      "entropy": 5.564316654205323,
      "epoch": 1.024275432795176,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899724247511299,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19270816296339036,
      "num_tokens": 22837411.0,
      "step": 13165
    },
    {
      "entropy": 5.487604475021362,
      "epoch": 1.0246644621668937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489964241256021,
      "loss": 5.0741,
      "mean_token_accuracy": 0.19337802827358247,
      "num_tokens": 22846265.0,
      "step": 13170
    },
    {
      "entropy": 5.484097814559936,
      "epoch": 1.0250534915386111,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004899560544991718,
      "loss": 5.0325,
      "mean_token_accuracy": 0.19119783490896225,
      "num_tokens": 22854489.0,
      "step": 13175
    },
    {
      "entropy": 5.589381361007691,
      "epoch": 1.0254425209103288,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004899478644807068,
      "loss": 5.2139,
      "mean_token_accuracy": 0.18560332357883452,
      "num_tokens": 22863269.0,
      "step": 13180
    },
    {
      "entropy": 5.554958152770996,
      "epoch": 1.0258315502820463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899396712007498,
      "loss": 5.1829,
      "mean_token_accuracy": 0.18641023337841034,
      "num_tokens": 22872304.0,
      "step": 13185
    },
    {
      "entropy": 5.488417053222657,
      "epoch": 1.0262205796537638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004899314746594256,
      "loss": 4.9906,
      "mean_token_accuracy": 0.19376779794692994,
      "num_tokens": 22881409.0,
      "step": 13190
    },
    {
      "entropy": 5.533855533599853,
      "epoch": 1.0266096090254815,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004899232748568584,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1938805639743805,
      "num_tokens": 22889560.0,
      "step": 13195
    },
    {
      "entropy": 5.505529689788818,
      "epoch": 1.026998638397199,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004899150717931724,
      "loss": 5.1683,
      "mean_token_accuracy": 0.1879460722208023,
      "num_tokens": 22897434.0,
      "step": 13200
    },
    {
      "entropy": 5.53561487197876,
      "epoch": 1.0273876677689167,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004899068654684924,
      "loss": 5.1345,
      "mean_token_accuracy": 0.18807875663042067,
      "num_tokens": 22906050.0,
      "step": 13205
    },
    {
      "entropy": 5.555013132095337,
      "epoch": 1.0277766971406341,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004898986558829428,
      "loss": 5.0929,
      "mean_token_accuracy": 0.18595127016305923,
      "num_tokens": 22914118.0,
      "step": 13210
    },
    {
      "entropy": 5.5879463195800785,
      "epoch": 1.0281657265123516,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004898904430366479,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19043147563934326,
      "num_tokens": 22922539.0,
      "step": 13215
    },
    {
      "entropy": 5.528032827377319,
      "epoch": 1.0285547558840693,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898822269297328,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19073048681020738,
      "num_tokens": 22932371.0,
      "step": 13220
    },
    {
      "entropy": 5.567050933837891,
      "epoch": 1.0289437852557868,
      "grad_norm": 1.125,
      "learning_rate": 0.0004898740075623218,
      "loss": 5.1584,
      "mean_token_accuracy": 0.18682138472795487,
      "num_tokens": 22940608.0,
      "step": 13225
    },
    {
      "entropy": 5.575891399383545,
      "epoch": 1.0293328146275045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004898657849345399,
      "loss": 5.1226,
      "mean_token_accuracy": 0.18368037194013595,
      "num_tokens": 22948905.0,
      "step": 13230
    },
    {
      "entropy": 5.4839098930358885,
      "epoch": 1.029721843999222,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004898575590465116,
      "loss": 5.092,
      "mean_token_accuracy": 0.18717192560434343,
      "num_tokens": 22957218.0,
      "step": 13235
    },
    {
      "entropy": 5.568029880523682,
      "epoch": 1.0301108733709394,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004898493298983619,
      "loss": 5.2425,
      "mean_token_accuracy": 0.18639528155326843,
      "num_tokens": 22965700.0,
      "step": 13240
    },
    {
      "entropy": 5.651806831359863,
      "epoch": 1.0304999027426571,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004898410974902155,
      "loss": 5.1974,
      "mean_token_accuracy": 0.18587371557950974,
      "num_tokens": 22973996.0,
      "step": 13245
    },
    {
      "entropy": 5.635276412963867,
      "epoch": 1.0308889321143746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004898328618221974,
      "loss": 5.1744,
      "mean_token_accuracy": 0.18107367753982545,
      "num_tokens": 22983142.0,
      "step": 13250
    },
    {
      "entropy": 5.552687788009644,
      "epoch": 1.0312779614860923,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004898246228944327,
      "loss": 5.0939,
      "mean_token_accuracy": 0.19192184507846832,
      "num_tokens": 22991263.0,
      "step": 13255
    },
    {
      "entropy": 5.585746097564697,
      "epoch": 1.0316669908578098,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898163807070461,
      "loss": 5.2065,
      "mean_token_accuracy": 0.18342706114053725,
      "num_tokens": 23000068.0,
      "step": 13260
    },
    {
      "entropy": 5.575736856460571,
      "epoch": 1.0320560202295272,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004898081352601629,
      "loss": 5.1071,
      "mean_token_accuracy": 0.18620418459177018,
      "num_tokens": 23007935.0,
      "step": 13265
    },
    {
      "entropy": 5.55137186050415,
      "epoch": 1.032445049601245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004897998865539082,
      "loss": 5.197,
      "mean_token_accuracy": 0.17447893023490907,
      "num_tokens": 23017462.0,
      "step": 13270
    },
    {
      "entropy": 5.615472841262817,
      "epoch": 1.0328340789729624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897916345884069,
      "loss": 5.1637,
      "mean_token_accuracy": 0.18629585653543473,
      "num_tokens": 23026681.0,
      "step": 13275
    },
    {
      "entropy": 5.553278541564941,
      "epoch": 1.0332231083446801,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897833793637847,
      "loss": 5.1462,
      "mean_token_accuracy": 0.18213918358087539,
      "num_tokens": 23035720.0,
      "step": 13280
    },
    {
      "entropy": 5.614792203903198,
      "epoch": 1.0336121377163976,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004897751208801665,
      "loss": 5.1445,
      "mean_token_accuracy": 0.18943334519863128,
      "num_tokens": 23044491.0,
      "step": 13285
    },
    {
      "entropy": 5.595033931732178,
      "epoch": 1.034001167088115,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897668591376777,
      "loss": 5.1348,
      "mean_token_accuracy": 0.18924394696950914,
      "num_tokens": 23054034.0,
      "step": 13290
    },
    {
      "entropy": 5.535338497161865,
      "epoch": 1.0343901964598328,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897585941364436,
      "loss": 5.1372,
      "mean_token_accuracy": 0.18637705743312835,
      "num_tokens": 23062478.0,
      "step": 13295
    },
    {
      "entropy": 5.552330541610718,
      "epoch": 1.0347792258315502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004897503258765896,
      "loss": 5.1207,
      "mean_token_accuracy": 0.1795137733221054,
      "num_tokens": 23070776.0,
      "step": 13300
    },
    {
      "entropy": 5.599796009063721,
      "epoch": 1.035168255203268,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004897420543582413,
      "loss": 5.1903,
      "mean_token_accuracy": 0.18395274579524995,
      "num_tokens": 23078459.0,
      "step": 13305
    },
    {
      "entropy": 5.561110877990723,
      "epoch": 1.0355572845749854,
      "grad_norm": 1.015625,
      "learning_rate": 0.000489733779581524,
      "loss": 5.1553,
      "mean_token_accuracy": 0.1920012965798378,
      "num_tokens": 23087113.0,
      "step": 13310
    },
    {
      "entropy": 5.516515684127808,
      "epoch": 1.035946313946703,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004897255015465635,
      "loss": 5.0795,
      "mean_token_accuracy": 0.18899349868297577,
      "num_tokens": 23095989.0,
      "step": 13315
    },
    {
      "entropy": 5.5466451168060305,
      "epoch": 1.0363353433184206,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004897172202534853,
      "loss": 5.1164,
      "mean_token_accuracy": 0.1875885471701622,
      "num_tokens": 23104578.0,
      "step": 13320
    },
    {
      "entropy": 5.566480922698974,
      "epoch": 1.036724372690138,
      "grad_norm": 1.03125,
      "learning_rate": 0.000489708935702415,
      "loss": 5.1606,
      "mean_token_accuracy": 0.18513909429311753,
      "num_tokens": 23114390.0,
      "step": 13325
    },
    {
      "entropy": 5.627627372741699,
      "epoch": 1.0371134020618558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004897006478934784,
      "loss": 5.0644,
      "mean_token_accuracy": 0.19613390415906906,
      "num_tokens": 23122411.0,
      "step": 13330
    },
    {
      "entropy": 5.494841766357422,
      "epoch": 1.0375024314335732,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004896923568268011,
      "loss": 4.9954,
      "mean_token_accuracy": 0.19336289316415786,
      "num_tokens": 23130801.0,
      "step": 13335
    },
    {
      "entropy": 5.517170667648315,
      "epoch": 1.0378914608052907,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000489684062502509,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19381724894046784,
      "num_tokens": 23138578.0,
      "step": 13340
    },
    {
      "entropy": 5.54047737121582,
      "epoch": 1.0382804901770084,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896757649207281,
      "loss": 5.1189,
      "mean_token_accuracy": 0.18910958617925644,
      "num_tokens": 23147135.0,
      "step": 13345
    },
    {
      "entropy": 5.634407567977905,
      "epoch": 1.0386695195487259,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000489667464081584,
      "loss": 5.2145,
      "mean_token_accuracy": 0.1848840057849884,
      "num_tokens": 23155791.0,
      "step": 13350
    },
    {
      "entropy": 5.6059342384338375,
      "epoch": 1.0390585489204436,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896591599852029,
      "loss": 5.069,
      "mean_token_accuracy": 0.19656053185462952,
      "num_tokens": 23163674.0,
      "step": 13355
    },
    {
      "entropy": 5.565481901168823,
      "epoch": 1.039447578292161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896508526317107,
      "loss": 5.1407,
      "mean_token_accuracy": 0.1826925128698349,
      "num_tokens": 23171931.0,
      "step": 13360
    },
    {
      "entropy": 5.570920944213867,
      "epoch": 1.0398366076638785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004896425420212334,
      "loss": 5.1467,
      "mean_token_accuracy": 0.1859249547123909,
      "num_tokens": 23180195.0,
      "step": 13365
    },
    {
      "entropy": 5.546844530105591,
      "epoch": 1.0402256370355962,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004896342281538973,
      "loss": 5.0744,
      "mean_token_accuracy": 0.1921488031744957,
      "num_tokens": 23189486.0,
      "step": 13370
    },
    {
      "entropy": 5.592347621917725,
      "epoch": 1.0406146664073137,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896259110298283,
      "loss": 5.2049,
      "mean_token_accuracy": 0.1832217127084732,
      "num_tokens": 23198055.0,
      "step": 13375
    },
    {
      "entropy": 5.546463441848755,
      "epoch": 1.0410036957790314,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004896175906491528,
      "loss": 5.0699,
      "mean_token_accuracy": 0.1903143882751465,
      "num_tokens": 23206092.0,
      "step": 13380
    },
    {
      "entropy": 5.556701612472534,
      "epoch": 1.0413927251507489,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896092670119968,
      "loss": 5.157,
      "mean_token_accuracy": 0.19196297228336334,
      "num_tokens": 23214820.0,
      "step": 13385
    },
    {
      "entropy": 5.504127407073975,
      "epoch": 1.0417817545224664,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004896009401184869,
      "loss": 5.1056,
      "mean_token_accuracy": 0.19079136550426484,
      "num_tokens": 23224033.0,
      "step": 13390
    },
    {
      "entropy": 5.526195096969604,
      "epoch": 1.042170783894184,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895926099687493,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1947038471698761,
      "num_tokens": 23233074.0,
      "step": 13395
    },
    {
      "entropy": 5.5074738502502445,
      "epoch": 1.0425598132659015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004895842765629104,
      "loss": 5.0756,
      "mean_token_accuracy": 0.18704090714454652,
      "num_tokens": 23241167.0,
      "step": 13400
    },
    {
      "entropy": 5.622848653793335,
      "epoch": 1.0429488426376192,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004895759399010966,
      "loss": 5.083,
      "mean_token_accuracy": 0.1889680355787277,
      "num_tokens": 23249601.0,
      "step": 13405
    },
    {
      "entropy": 5.594100666046143,
      "epoch": 1.0433378720093367,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004895675999834345,
      "loss": 5.1962,
      "mean_token_accuracy": 0.17248471677303315,
      "num_tokens": 23257868.0,
      "step": 13410
    },
    {
      "entropy": 5.5872410297393795,
      "epoch": 1.0437269013810542,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895592568100505,
      "loss": 5.1573,
      "mean_token_accuracy": 0.18521177470684053,
      "num_tokens": 23266586.0,
      "step": 13415
    },
    {
      "entropy": 5.619315814971924,
      "epoch": 1.0441159307527719,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004895509103810714,
      "loss": 5.1367,
      "mean_token_accuracy": 0.1923282578587532,
      "num_tokens": 23275114.0,
      "step": 13420
    },
    {
      "entropy": 5.479888391494751,
      "epoch": 1.0445049601244893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004895425606966237,
      "loss": 5.04,
      "mean_token_accuracy": 0.19801546335220338,
      "num_tokens": 23283200.0,
      "step": 13425
    },
    {
      "entropy": 5.592350769042969,
      "epoch": 1.044893989496207,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000489534207756834,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1798463299870491,
      "num_tokens": 23291999.0,
      "step": 13430
    },
    {
      "entropy": 5.587844371795654,
      "epoch": 1.0452830188679245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004895258515618293,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1921411395072937,
      "num_tokens": 23300514.0,
      "step": 13435
    },
    {
      "entropy": 5.563119411468506,
      "epoch": 1.045672048239642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004895174921117362,
      "loss": 5.0675,
      "mean_token_accuracy": 0.19010502099990845,
      "num_tokens": 23309543.0,
      "step": 13440
    },
    {
      "entropy": 5.510892009735107,
      "epoch": 1.0460610776113597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004895091294066816,
      "loss": 5.1156,
      "mean_token_accuracy": 0.18795071989297868,
      "num_tokens": 23319002.0,
      "step": 13445
    },
    {
      "entropy": 5.5731898784637455,
      "epoch": 1.0464501069830772,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004895007634467924,
      "loss": 5.1561,
      "mean_token_accuracy": 0.18822090625762938,
      "num_tokens": 23326604.0,
      "step": 13450
    },
    {
      "entropy": 5.521999263763428,
      "epoch": 1.0468391363547949,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004894923942321955,
      "loss": 5.0985,
      "mean_token_accuracy": 0.19514717906713486,
      "num_tokens": 23335492.0,
      "step": 13455
    },
    {
      "entropy": 5.55545597076416,
      "epoch": 1.0472281657265123,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004894840217630179,
      "loss": 5.095,
      "mean_token_accuracy": 0.19582296311855316,
      "num_tokens": 23343693.0,
      "step": 13460
    },
    {
      "entropy": 5.528950548171997,
      "epoch": 1.0476171950982298,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894756460393868,
      "loss": 5.0873,
      "mean_token_accuracy": 0.19680141508579255,
      "num_tokens": 23352587.0,
      "step": 13465
    },
    {
      "entropy": 5.539797735214234,
      "epoch": 1.0480062244699475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489467267061429,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1911733165383339,
      "num_tokens": 23361293.0,
      "step": 13470
    },
    {
      "entropy": 5.57023777961731,
      "epoch": 1.048395253841665,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004894588848292718,
      "loss": 5.1533,
      "mean_token_accuracy": 0.18564065992832185,
      "num_tokens": 23369663.0,
      "step": 13475
    },
    {
      "entropy": 5.518763685226441,
      "epoch": 1.0487842832133827,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894504993430425,
      "loss": 5.0588,
      "mean_token_accuracy": 0.1874425008893013,
      "num_tokens": 23378224.0,
      "step": 13480
    },
    {
      "entropy": 5.604974555969238,
      "epoch": 1.0491733125851002,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000489442110602868,
      "loss": 5.2074,
      "mean_token_accuracy": 0.18353474587202073,
      "num_tokens": 23386693.0,
      "step": 13485
    },
    {
      "entropy": 5.592781591415405,
      "epoch": 1.0495623419568176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004894337186088759,
      "loss": 5.0846,
      "mean_token_accuracy": 0.19203248620033264,
      "num_tokens": 23395226.0,
      "step": 13490
    },
    {
      "entropy": 5.5541750431060795,
      "epoch": 1.0499513713285353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004894253233611934,
      "loss": 5.122,
      "mean_token_accuracy": 0.184508316218853,
      "num_tokens": 23403779.0,
      "step": 13495
    },
    {
      "entropy": 5.610698413848877,
      "epoch": 1.0503404007002528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489416924859948,
      "loss": 5.2214,
      "mean_token_accuracy": 0.18582794517278672,
      "num_tokens": 23411807.0,
      "step": 13500
    },
    {
      "entropy": 5.542945480346679,
      "epoch": 1.0507294300719705,
      "grad_norm": 1.1875,
      "learning_rate": 0.000489408523105267,
      "loss": 5.1,
      "mean_token_accuracy": 0.1886272355914116,
      "num_tokens": 23420323.0,
      "step": 13505
    },
    {
      "entropy": 5.5439781665802,
      "epoch": 1.051118459443688,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004894001180972779,
      "loss": 5.1022,
      "mean_token_accuracy": 0.1886004999279976,
      "num_tokens": 23429220.0,
      "step": 13510
    },
    {
      "entropy": 5.565813875198364,
      "epoch": 1.0515074888154055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004893917098361082,
      "loss": 5.1182,
      "mean_token_accuracy": 0.19521507024765014,
      "num_tokens": 23438471.0,
      "step": 13515
    },
    {
      "entropy": 5.59785566329956,
      "epoch": 1.0518965181871232,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004893832983218856,
      "loss": 5.1793,
      "mean_token_accuracy": 0.1917217805981636,
      "num_tokens": 23447504.0,
      "step": 13520
    },
    {
      "entropy": 5.457501459121704,
      "epoch": 1.0522855475588406,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004893748835547377,
      "loss": 4.9281,
      "mean_token_accuracy": 0.20070867389440536,
      "num_tokens": 23456065.0,
      "step": 13525
    },
    {
      "entropy": 5.53516902923584,
      "epoch": 1.0526745769305583,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004893664655347921,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19218259155750275,
      "num_tokens": 23463826.0,
      "step": 13530
    },
    {
      "entropy": 5.465136623382568,
      "epoch": 1.0530636063022758,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004893580442621767,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20289442092180252,
      "num_tokens": 23472092.0,
      "step": 13535
    },
    {
      "entropy": 5.569958972930908,
      "epoch": 1.0534526356739935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000489349619737019,
      "loss": 5.1565,
      "mean_token_accuracy": 0.18465171605348588,
      "num_tokens": 23480934.0,
      "step": 13540
    },
    {
      "entropy": 5.603713035583496,
      "epoch": 1.053841665045711,
      "grad_norm": 1.03125,
      "learning_rate": 0.000489341191959447,
      "loss": 5.1286,
      "mean_token_accuracy": 0.1867547392845154,
      "num_tokens": 23490187.0,
      "step": 13545
    },
    {
      "entropy": 5.572095727920532,
      "epoch": 1.0542306944174284,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004893327609295887,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1879362866282463,
      "num_tokens": 23498330.0,
      "step": 13550
    },
    {
      "entropy": 5.488249921798706,
      "epoch": 1.0546197237891461,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004893243266475719,
      "loss": 5.0194,
      "mean_token_accuracy": 0.18986970633268357,
      "num_tokens": 23506219.0,
      "step": 13555
    },
    {
      "entropy": 5.517959117889404,
      "epoch": 1.0550087531608636,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004893158891135245,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19314618557691574,
      "num_tokens": 23514585.0,
      "step": 13560
    },
    {
      "entropy": 5.586368560791016,
      "epoch": 1.055397782532581,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004893074483275746,
      "loss": 5.0591,
      "mean_token_accuracy": 0.19126506447792052,
      "num_tokens": 23522919.0,
      "step": 13565
    },
    {
      "entropy": 5.624224424362183,
      "epoch": 1.0557868119042988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892990042898503,
      "loss": 5.2578,
      "mean_token_accuracy": 0.1761441320180893,
      "num_tokens": 23531613.0,
      "step": 13570
    },
    {
      "entropy": 5.610091733932495,
      "epoch": 1.0561758412760163,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004892905570004798,
      "loss": 5.2124,
      "mean_token_accuracy": 0.18559568524360656,
      "num_tokens": 23539865.0,
      "step": 13575
    },
    {
      "entropy": 5.51715497970581,
      "epoch": 1.056564870647734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000489282106459591,
      "loss": 5.0452,
      "mean_token_accuracy": 0.19779579043388368,
      "num_tokens": 23548424.0,
      "step": 13580
    },
    {
      "entropy": 5.569211626052857,
      "epoch": 1.0569539000194514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892736526673124,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19448269307613372,
      "num_tokens": 23557199.0,
      "step": 13585
    },
    {
      "entropy": 5.592437887191773,
      "epoch": 1.0573429293911691,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004892651956237721,
      "loss": 5.2468,
      "mean_token_accuracy": 0.17913527339696883,
      "num_tokens": 23565979.0,
      "step": 13590
    },
    {
      "entropy": 5.5637218952178955,
      "epoch": 1.0577319587628866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892567353290986,
      "loss": 5.1103,
      "mean_token_accuracy": 0.18697206526994706,
      "num_tokens": 23575405.0,
      "step": 13595
    },
    {
      "entropy": 5.659807014465332,
      "epoch": 1.058120988134604,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004892482717834201,
      "loss": 5.2517,
      "mean_token_accuracy": 0.1756979763507843,
      "num_tokens": 23584589.0,
      "step": 13600
    },
    {
      "entropy": 5.548164319992066,
      "epoch": 1.0585100175063218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004892398049868651,
      "loss": 5.1075,
      "mean_token_accuracy": 0.18530123978853225,
      "num_tokens": 23593079.0,
      "step": 13605
    },
    {
      "entropy": 5.566551303863525,
      "epoch": 1.0588990468780393,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000489231334939562,
      "loss": 5.186,
      "mean_token_accuracy": 0.19026433676481247,
      "num_tokens": 23602010.0,
      "step": 13610
    },
    {
      "entropy": 5.630366563796997,
      "epoch": 1.059288076249757,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004892228616416395,
      "loss": 5.1449,
      "mean_token_accuracy": 0.188672237098217,
      "num_tokens": 23610707.0,
      "step": 13615
    },
    {
      "entropy": 5.526639747619629,
      "epoch": 1.0596771056214744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892143850932259,
      "loss": 5.1019,
      "mean_token_accuracy": 0.18711029440164567,
      "num_tokens": 23619579.0,
      "step": 13620
    },
    {
      "entropy": 5.530226469039917,
      "epoch": 1.060066134993192,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048920590529445,
      "loss": 5.079,
      "mean_token_accuracy": 0.2033025175333023,
      "num_tokens": 23628407.0,
      "step": 13625
    },
    {
      "entropy": 5.4833169937133786,
      "epoch": 1.0604551643649096,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004891974222454406,
      "loss": 5.0316,
      "mean_token_accuracy": 0.19568670839071273,
      "num_tokens": 23637094.0,
      "step": 13630
    },
    {
      "entropy": 5.513839054107666,
      "epoch": 1.060844193736627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004891889359463261,
      "loss": 5.0939,
      "mean_token_accuracy": 0.19252139925956727,
      "num_tokens": 23645823.0,
      "step": 13635
    },
    {
      "entropy": 5.625003290176392,
      "epoch": 1.0612332231083448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891804463972354,
      "loss": 5.2009,
      "mean_token_accuracy": 0.18133169412612915,
      "num_tokens": 23655027.0,
      "step": 13640
    },
    {
      "entropy": 5.660948419570923,
      "epoch": 1.0616222524800623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004891719535982974,
      "loss": 5.2634,
      "mean_token_accuracy": 0.17668553143739701,
      "num_tokens": 23663840.0,
      "step": 13645
    },
    {
      "entropy": 5.621017217636108,
      "epoch": 1.0620112818517797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891634575496408,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1838197663426399,
      "num_tokens": 23672795.0,
      "step": 13650
    },
    {
      "entropy": 5.5811532497406,
      "epoch": 1.0624003112234974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004891549582513946,
      "loss": 5.1358,
      "mean_token_accuracy": 0.18807750940322876,
      "num_tokens": 23681833.0,
      "step": 13655
    },
    {
      "entropy": 5.573037958145141,
      "epoch": 1.062789340595215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000489146455703688,
      "loss": 5.1039,
      "mean_token_accuracy": 0.19066709131002427,
      "num_tokens": 23689939.0,
      "step": 13660
    },
    {
      "entropy": 5.523500299453735,
      "epoch": 1.0631783699669326,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004891379499066495,
      "loss": 5.1604,
      "mean_token_accuracy": 0.19016223102808,
      "num_tokens": 23698535.0,
      "step": 13665
    },
    {
      "entropy": 5.504574632644653,
      "epoch": 1.06356739933865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891294408604087,
      "loss": 5.078,
      "mean_token_accuracy": 0.19258528500795363,
      "num_tokens": 23707749.0,
      "step": 13670
    },
    {
      "entropy": 5.580240201950073,
      "epoch": 1.0639564287103676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004891209285650942,
      "loss": 5.2118,
      "mean_token_accuracy": 0.19406515061855317,
      "num_tokens": 23716503.0,
      "step": 13675
    },
    {
      "entropy": 5.62525053024292,
      "epoch": 1.0643454580820852,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004891124130208355,
      "loss": 5.1755,
      "mean_token_accuracy": 0.1868963807821274,
      "num_tokens": 23724882.0,
      "step": 13680
    },
    {
      "entropy": 5.491151809692383,
      "epoch": 1.0647344874538027,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004891038942277618,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1910150021314621,
      "num_tokens": 23733680.0,
      "step": 13685
    },
    {
      "entropy": 5.535544586181641,
      "epoch": 1.0651235168255204,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004890953721860022,
      "loss": 5.1083,
      "mean_token_accuracy": 0.1963130235671997,
      "num_tokens": 23742066.0,
      "step": 13690
    },
    {
      "entropy": 5.539183616638184,
      "epoch": 1.065512546197238,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004890868468956862,
      "loss": 5.0252,
      "mean_token_accuracy": 0.19888394325971603,
      "num_tokens": 23750558.0,
      "step": 13695
    },
    {
      "entropy": 5.5164039611816404,
      "epoch": 1.0659015755689554,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489078318356943,
      "loss": 5.0895,
      "mean_token_accuracy": 0.18827929198741913,
      "num_tokens": 23759120.0,
      "step": 13700
    },
    {
      "entropy": 5.5378562927246096,
      "epoch": 1.066290604940673,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004890697865699021,
      "loss": 5.1308,
      "mean_token_accuracy": 0.18907262831926347,
      "num_tokens": 23768301.0,
      "step": 13705
    },
    {
      "entropy": 5.5610710144042965,
      "epoch": 1.0666796343123905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004890612515346929,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19625268876552582,
      "num_tokens": 23776155.0,
      "step": 13710
    },
    {
      "entropy": 5.56669864654541,
      "epoch": 1.0670686636841082,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004890527132514448,
      "loss": 5.1716,
      "mean_token_accuracy": 0.18320638835430145,
      "num_tokens": 23785363.0,
      "step": 13715
    },
    {
      "entropy": 5.537170839309693,
      "epoch": 1.0674576930558257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004890441717202876,
      "loss": 5.1214,
      "mean_token_accuracy": 0.1906835541129112,
      "num_tokens": 23793817.0,
      "step": 13720
    },
    {
      "entropy": 5.515767574310303,
      "epoch": 1.0678467224275432,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004890356269413507,
      "loss": 5.1712,
      "mean_token_accuracy": 0.18317762315273284,
      "num_tokens": 23802629.0,
      "step": 13725
    },
    {
      "entropy": 5.5109014987945555,
      "epoch": 1.068235751799261,
      "grad_norm": 1.140625,
      "learning_rate": 0.000489027078914764,
      "loss": 4.9687,
      "mean_token_accuracy": 0.20254837423563005,
      "num_tokens": 23811178.0,
      "step": 13730
    },
    {
      "entropy": 5.598137283325196,
      "epoch": 1.0686247811709784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004890185276406569,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18314220607280732,
      "num_tokens": 23819685.0,
      "step": 13735
    },
    {
      "entropy": 5.455999565124512,
      "epoch": 1.069013810542696,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890099731191592,
      "loss": 5.0632,
      "mean_token_accuracy": 0.18806861340999603,
      "num_tokens": 23828733.0,
      "step": 13740
    },
    {
      "entropy": 5.620703268051147,
      "epoch": 1.0694028399144135,
      "grad_norm": 1.15625,
      "learning_rate": 0.000489001415350401,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1827014997601509,
      "num_tokens": 23837095.0,
      "step": 13745
    },
    {
      "entropy": 5.614408349990844,
      "epoch": 1.069791869286131,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004889928543345118,
      "loss": 5.1825,
      "mean_token_accuracy": 0.18442823737859726,
      "num_tokens": 23846637.0,
      "step": 13750
    },
    {
      "entropy": 5.6199158191680905,
      "epoch": 1.0701808986578487,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004889842900716217,
      "loss": 5.2468,
      "mean_token_accuracy": 0.17961147725582122,
      "num_tokens": 23856042.0,
      "step": 13755
    },
    {
      "entropy": 5.608266592025757,
      "epoch": 1.0705699280295662,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889757225618606,
      "loss": 5.2012,
      "mean_token_accuracy": 0.17659927159547806,
      "num_tokens": 23865367.0,
      "step": 13760
    },
    {
      "entropy": 5.585307598114014,
      "epoch": 1.0709589574012839,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004889671518053584,
      "loss": 5.1979,
      "mean_token_accuracy": 0.1872079610824585,
      "num_tokens": 23874740.0,
      "step": 13765
    },
    {
      "entropy": 5.593739986419678,
      "epoch": 1.0713479867730014,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004889585778022453,
      "loss": 5.1353,
      "mean_token_accuracy": 0.1876058906316757,
      "num_tokens": 23884229.0,
      "step": 13770
    },
    {
      "entropy": 5.524658918380737,
      "epoch": 1.0717370161447188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004889500005526514,
      "loss": 5.0728,
      "mean_token_accuracy": 0.1901668354868889,
      "num_tokens": 23892149.0,
      "step": 13775
    },
    {
      "entropy": 5.655459213256836,
      "epoch": 1.0721260455164365,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004889414200567067,
      "loss": 5.2703,
      "mean_token_accuracy": 0.17439593970775605,
      "num_tokens": 23901802.0,
      "step": 13780
    },
    {
      "entropy": 5.634878635406494,
      "epoch": 1.072515074888154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004889328363145415,
      "loss": 5.13,
      "mean_token_accuracy": 0.18976068645715713,
      "num_tokens": 23910268.0,
      "step": 13785
    },
    {
      "entropy": 5.453121614456177,
      "epoch": 1.0729041042598717,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889242493262859,
      "loss": 4.9909,
      "mean_token_accuracy": 0.19566470086574556,
      "num_tokens": 23918351.0,
      "step": 13790
    },
    {
      "entropy": 5.50495867729187,
      "epoch": 1.0732931336315892,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004889156590920704,
      "loss": 5.0347,
      "mean_token_accuracy": 0.19083705693483352,
      "num_tokens": 23927367.0,
      "step": 13795
    },
    {
      "entropy": 5.543604135513306,
      "epoch": 1.0736821630033067,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004889070656120253,
      "loss": 5.1255,
      "mean_token_accuracy": 0.1959057942032814,
      "num_tokens": 23936216.0,
      "step": 13800
    },
    {
      "entropy": 5.641427946090698,
      "epoch": 1.0740711923750244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004888984688862809,
      "loss": 5.1849,
      "mean_token_accuracy": 0.18578004240989685,
      "num_tokens": 23944344.0,
      "step": 13805
    },
    {
      "entropy": 5.599825239181518,
      "epoch": 1.0744602217467418,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004888898689149677,
      "loss": 5.2214,
      "mean_token_accuracy": 0.17829657196998597,
      "num_tokens": 23953390.0,
      "step": 13810
    },
    {
      "entropy": 5.661473798751831,
      "epoch": 1.0748492511184595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004888812656982162,
      "loss": 5.1981,
      "mean_token_accuracy": 0.18864652514457703,
      "num_tokens": 23962230.0,
      "step": 13815
    },
    {
      "entropy": 5.533057785034179,
      "epoch": 1.075238280490177,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004888726592361569,
      "loss": 5.1186,
      "mean_token_accuracy": 0.18507571816444396,
      "num_tokens": 23971627.0,
      "step": 13820
    },
    {
      "entropy": 5.5385621070861815,
      "epoch": 1.0756273098618945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004888640495289204,
      "loss": 5.0874,
      "mean_token_accuracy": 0.19487264901399612,
      "num_tokens": 23980222.0,
      "step": 13825
    },
    {
      "entropy": 5.524286222457886,
      "epoch": 1.0760163392336122,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888554365766374,
      "loss": 5.1698,
      "mean_token_accuracy": 0.18793643563985823,
      "num_tokens": 23988803.0,
      "step": 13830
    },
    {
      "entropy": 5.5675934791564945,
      "epoch": 1.0764053686053296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004888468203794385,
      "loss": 5.1594,
      "mean_token_accuracy": 0.18618838489055634,
      "num_tokens": 23997664.0,
      "step": 13835
    },
    {
      "entropy": 5.565265798568726,
      "epoch": 1.0767943979770473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004888382009374545,
      "loss": 5.1289,
      "mean_token_accuracy": 0.188122895359993,
      "num_tokens": 24006641.0,
      "step": 13840
    },
    {
      "entropy": 5.515845823287964,
      "epoch": 1.0771834273487648,
      "grad_norm": 1.125,
      "learning_rate": 0.0004888295782508161,
      "loss": 5.0817,
      "mean_token_accuracy": 0.18674434572458268,
      "num_tokens": 24014852.0,
      "step": 13845
    },
    {
      "entropy": 5.576596117019653,
      "epoch": 1.0775724567204823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004888209523196542,
      "loss": 5.1234,
      "mean_token_accuracy": 0.1953816831111908,
      "num_tokens": 24023441.0,
      "step": 13850
    },
    {
      "entropy": 5.6271281242370605,
      "epoch": 1.0779614860922,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004888123231440997,
      "loss": 5.1274,
      "mean_token_accuracy": 0.18886868059635162,
      "num_tokens": 24032120.0,
      "step": 13855
    },
    {
      "entropy": 5.587115049362183,
      "epoch": 1.0783505154639175,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888036907242834,
      "loss": 5.1776,
      "mean_token_accuracy": 0.1870588704943657,
      "num_tokens": 24040264.0,
      "step": 13860
    },
    {
      "entropy": 5.598802852630615,
      "epoch": 1.0787395448356352,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004887950550603366,
      "loss": 5.2165,
      "mean_token_accuracy": 0.185377599298954,
      "num_tokens": 24049194.0,
      "step": 13865
    },
    {
      "entropy": 5.551940250396728,
      "epoch": 1.0791285742073526,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004887864161523901,
      "loss": 5.1055,
      "mean_token_accuracy": 0.18662161529064178,
      "num_tokens": 24058004.0,
      "step": 13870
    },
    {
      "entropy": 5.507250070571899,
      "epoch": 1.0795176035790701,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887777740005749,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1873527228832245,
      "num_tokens": 24066871.0,
      "step": 13875
    },
    {
      "entropy": 5.5591918468475345,
      "epoch": 1.0799066329507878,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004887691286050224,
      "loss": 5.139,
      "mean_token_accuracy": 0.18786173164844513,
      "num_tokens": 24075409.0,
      "step": 13880
    },
    {
      "entropy": 5.596242380142212,
      "epoch": 1.0802956623225053,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004887604799658635,
      "loss": 5.0979,
      "mean_token_accuracy": 0.1871548518538475,
      "num_tokens": 24083983.0,
      "step": 13885
    },
    {
      "entropy": 5.574283313751221,
      "epoch": 1.080684691694223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004887518280832295,
      "loss": 5.1316,
      "mean_token_accuracy": 0.19244279712438583,
      "num_tokens": 24092225.0,
      "step": 13890
    },
    {
      "entropy": 5.493574380874634,
      "epoch": 1.0810737210659405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004887431729572519,
      "loss": 5.0436,
      "mean_token_accuracy": 0.19515586793422698,
      "num_tokens": 24100045.0,
      "step": 13895
    },
    {
      "entropy": 5.510802173614502,
      "epoch": 1.081462750437658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004887345145880617,
      "loss": 5.1184,
      "mean_token_accuracy": 0.19693547934293748,
      "num_tokens": 24108540.0,
      "step": 13900
    },
    {
      "entropy": 5.572795391082764,
      "epoch": 1.0818517798093756,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004887258529757904,
      "loss": 5.1684,
      "mean_token_accuracy": 0.18873950839042664,
      "num_tokens": 24117819.0,
      "step": 13905
    },
    {
      "entropy": 5.551358270645141,
      "epoch": 1.082240809181093,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004887171881205696,
      "loss": 4.9949,
      "mean_token_accuracy": 0.1976578488945961,
      "num_tokens": 24127000.0,
      "step": 13910
    },
    {
      "entropy": 5.566794490814209,
      "epoch": 1.0826298385528108,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004887085200225306,
      "loss": 5.1385,
      "mean_token_accuracy": 0.18214566856622696,
      "num_tokens": 24135404.0,
      "step": 13915
    },
    {
      "entropy": 5.476736545562744,
      "epoch": 1.0830188679245283,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004886998486818049,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20143810361623765,
      "num_tokens": 24143944.0,
      "step": 13920
    },
    {
      "entropy": 5.548741817474365,
      "epoch": 1.083407897296246,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004886911740985242,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1896209940314293,
      "num_tokens": 24153097.0,
      "step": 13925
    },
    {
      "entropy": 5.602055263519287,
      "epoch": 1.0837969266679635,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048868249627282,
      "loss": 5.212,
      "mean_token_accuracy": 0.18411155343055724,
      "num_tokens": 24162499.0,
      "step": 13930
    },
    {
      "entropy": 5.536603927612305,
      "epoch": 1.084185956039681,
      "grad_norm": 1.09375,
      "learning_rate": 0.000488673815204824,
      "loss": 5.0545,
      "mean_token_accuracy": 0.19690704941749573,
      "num_tokens": 24170543.0,
      "step": 13935
    },
    {
      "entropy": 5.513830375671387,
      "epoch": 1.0845749854113986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004886651308946681,
      "loss": 5.1151,
      "mean_token_accuracy": 0.18416737914085388,
      "num_tokens": 24180871.0,
      "step": 13940
    },
    {
      "entropy": 5.486499404907226,
      "epoch": 1.084964014783116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004886564433424839,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19448592513799667,
      "num_tokens": 24189075.0,
      "step": 13945
    },
    {
      "entropy": 5.563876390457153,
      "epoch": 1.0853530441548336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886477525484032,
      "loss": 5.1636,
      "mean_token_accuracy": 0.18980545699596404,
      "num_tokens": 24197686.0,
      "step": 13950
    },
    {
      "entropy": 5.503100204467773,
      "epoch": 1.0857420735265513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886390585125579,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19427724182605743,
      "num_tokens": 24207093.0,
      "step": 13955
    },
    {
      "entropy": 5.632514238357544,
      "epoch": 1.0861311028982688,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886303612350799,
      "loss": 5.1577,
      "mean_token_accuracy": 0.18540811389684678,
      "num_tokens": 24216447.0,
      "step": 13960
    },
    {
      "entropy": 5.621713018417358,
      "epoch": 1.0865201322699864,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004886216607161013,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1832515463232994,
      "num_tokens": 24225041.0,
      "step": 13965
    },
    {
      "entropy": 5.575142574310303,
      "epoch": 1.086909161641704,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004886129569557538,
      "loss": 5.131,
      "mean_token_accuracy": 0.1892656996846199,
      "num_tokens": 24233686.0,
      "step": 13970
    },
    {
      "entropy": 5.596088123321533,
      "epoch": 1.0872981910134216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004886042499541699,
      "loss": 5.1546,
      "mean_token_accuracy": 0.19379397928714753,
      "num_tokens": 24242553.0,
      "step": 13975
    },
    {
      "entropy": 5.501001977920533,
      "epoch": 1.087687220385139,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885955397114813,
      "loss": 5.0751,
      "mean_token_accuracy": 0.19152675569057465,
      "num_tokens": 24251658.0,
      "step": 13980
    },
    {
      "entropy": 5.568095397949219,
      "epoch": 1.0880762497568566,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004885868262278205,
      "loss": 5.1516,
      "mean_token_accuracy": 0.18734297752380372,
      "num_tokens": 24260397.0,
      "step": 13985
    },
    {
      "entropy": 5.5260601997375485,
      "epoch": 1.0884652791285743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885781095033195,
      "loss": 5.1045,
      "mean_token_accuracy": 0.18965977281332017,
      "num_tokens": 24268789.0,
      "step": 13990
    },
    {
      "entropy": 5.556467580795288,
      "epoch": 1.0888543085002917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885693895381108,
      "loss": 5.1927,
      "mean_token_accuracy": 0.18240671157836913,
      "num_tokens": 24278050.0,
      "step": 13995
    },
    {
      "entropy": 5.592872095108032,
      "epoch": 1.0892433378720092,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885606663323263,
      "loss": 5.1469,
      "mean_token_accuracy": 0.18555470257997514,
      "num_tokens": 24286477.0,
      "step": 14000
    },
    {
      "entropy": 5.467622137069702,
      "epoch": 1.089632367243727,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004885519398860987,
      "loss": 5.0237,
      "mean_token_accuracy": 0.196824412047863,
      "num_tokens": 24295704.0,
      "step": 14005
    },
    {
      "entropy": 5.471104288101197,
      "epoch": 1.0900213966154444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885432101995604,
      "loss": 5.0704,
      "mean_token_accuracy": 0.19854490011930465,
      "num_tokens": 24304132.0,
      "step": 14010
    },
    {
      "entropy": 5.576385450363159,
      "epoch": 1.090410425987162,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885344772728436,
      "loss": 5.154,
      "mean_token_accuracy": 0.1883609801530838,
      "num_tokens": 24312864.0,
      "step": 14015
    },
    {
      "entropy": 5.573934173583984,
      "epoch": 1.0907994553588796,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885257411060812,
      "loss": 5.0805,
      "mean_token_accuracy": 0.1923179268836975,
      "num_tokens": 24321745.0,
      "step": 14020
    },
    {
      "entropy": 5.512551212310791,
      "epoch": 1.0911884847305973,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885170016994053,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19153497815132142,
      "num_tokens": 24331061.0,
      "step": 14025
    },
    {
      "entropy": 5.445580005645752,
      "epoch": 1.0915775141023147,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885082590529489,
      "loss": 5.0929,
      "mean_token_accuracy": 0.18691821843385698,
      "num_tokens": 24339935.0,
      "step": 14030
    },
    {
      "entropy": 5.55517988204956,
      "epoch": 1.0919665434740322,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004884995131668445,
      "loss": 5.1193,
      "mean_token_accuracy": 0.1883410930633545,
      "num_tokens": 24348876.0,
      "step": 14035
    },
    {
      "entropy": 5.583040714263916,
      "epoch": 1.09235557284575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884907640412248,
      "loss": 5.1443,
      "mean_token_accuracy": 0.1901978150010109,
      "num_tokens": 24357094.0,
      "step": 14040
    },
    {
      "entropy": 5.5244217872619625,
      "epoch": 1.0927446022174674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884820116762225,
      "loss": 5.0987,
      "mean_token_accuracy": 0.18860751390457153,
      "num_tokens": 24366320.0,
      "step": 14045
    },
    {
      "entropy": 5.492645120620727,
      "epoch": 1.093133631589185,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004884732560719706,
      "loss": 5.0083,
      "mean_token_accuracy": 0.19360964596271515,
      "num_tokens": 24374390.0,
      "step": 14050
    },
    {
      "entropy": 5.51938214302063,
      "epoch": 1.0935226609609026,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004884644972286017,
      "loss": 5.1136,
      "mean_token_accuracy": 0.1967022016644478,
      "num_tokens": 24382804.0,
      "step": 14055
    },
    {
      "entropy": 5.572849273681641,
      "epoch": 1.09391169033262,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004884557351462488,
      "loss": 5.1141,
      "mean_token_accuracy": 0.1901245504617691,
      "num_tokens": 24391352.0,
      "step": 14060
    },
    {
      "entropy": 5.594938659667969,
      "epoch": 1.0943007197043377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884469698250449,
      "loss": 5.1714,
      "mean_token_accuracy": 0.18564601242542267,
      "num_tokens": 24400348.0,
      "step": 14065
    },
    {
      "entropy": 5.51166615486145,
      "epoch": 1.0946897490760552,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000488438201265123,
      "loss": 5.0138,
      "mean_token_accuracy": 0.19799385517835616,
      "num_tokens": 24409628.0,
      "step": 14070
    },
    {
      "entropy": 5.525060606002808,
      "epoch": 1.095078778447773,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004884294294666161,
      "loss": 5.1378,
      "mean_token_accuracy": 0.1854052498936653,
      "num_tokens": 24418379.0,
      "step": 14075
    },
    {
      "entropy": 5.642554044723511,
      "epoch": 1.0954678078194904,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004884206544296573,
      "loss": 5.2442,
      "mean_token_accuracy": 0.1798141285777092,
      "num_tokens": 24427719.0,
      "step": 14080
    },
    {
      "entropy": 5.6166298389434814,
      "epoch": 1.0958568371912079,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004884118761543797,
      "loss": 5.2049,
      "mean_token_accuracy": 0.18280501216650008,
      "num_tokens": 24435938.0,
      "step": 14085
    },
    {
      "entropy": 5.491099643707275,
      "epoch": 1.0962458665629256,
      "grad_norm": 1.125,
      "learning_rate": 0.0004884030946409167,
      "loss": 5.0824,
      "mean_token_accuracy": 0.19653500914573668,
      "num_tokens": 24444286.0,
      "step": 14090
    },
    {
      "entropy": 5.486567497253418,
      "epoch": 1.096634895934643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004883943098894013,
      "loss": 5.0241,
      "mean_token_accuracy": 0.19520906656980513,
      "num_tokens": 24452546.0,
      "step": 14095
    },
    {
      "entropy": 5.580649185180664,
      "epoch": 1.0970239253063607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000488385521899967,
      "loss": 5.1632,
      "mean_token_accuracy": 0.1858903467655182,
      "num_tokens": 24460765.0,
      "step": 14100
    },
    {
      "entropy": 5.5441375255584715,
      "epoch": 1.0974129546780782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000488376730672747,
      "loss": 5.0538,
      "mean_token_accuracy": 0.1881365016102791,
      "num_tokens": 24470018.0,
      "step": 14105
    },
    {
      "entropy": 5.504387903213501,
      "epoch": 1.0978019840497957,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004883679362078746,
      "loss": 4.9941,
      "mean_token_accuracy": 0.18995364010334015,
      "num_tokens": 24478882.0,
      "step": 14110
    },
    {
      "entropy": 5.570180225372314,
      "epoch": 1.0981910134215134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004883591385054836,
      "loss": 5.1598,
      "mean_token_accuracy": 0.18365177363157273,
      "num_tokens": 24487801.0,
      "step": 14115
    },
    {
      "entropy": 5.543331575393677,
      "epoch": 1.0985800427932308,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004883503375657072,
      "loss": 5.0985,
      "mean_token_accuracy": 0.1899980828166008,
      "num_tokens": 24496082.0,
      "step": 14120
    },
    {
      "entropy": 5.544586086273194,
      "epoch": 1.0989690721649485,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004883415333886789,
      "loss": 5.0392,
      "mean_token_accuracy": 0.1970685601234436,
      "num_tokens": 24504664.0,
      "step": 14125
    },
    {
      "entropy": 5.519941234588623,
      "epoch": 1.099358101536666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004883327259745325,
      "loss": 5.085,
      "mean_token_accuracy": 0.1909290924668312,
      "num_tokens": 24512797.0,
      "step": 14130
    },
    {
      "entropy": 5.489930772781372,
      "epoch": 1.0997471309083835,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004883239153234015,
      "loss": 5.058,
      "mean_token_accuracy": 0.1956195667386055,
      "num_tokens": 24520941.0,
      "step": 14135
    },
    {
      "entropy": 5.5680016040802,
      "epoch": 1.1001361602801012,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004883151014354197,
      "loss": 5.1502,
      "mean_token_accuracy": 0.18419487178325653,
      "num_tokens": 24530087.0,
      "step": 14140
    },
    {
      "entropy": 5.575088977813721,
      "epoch": 1.1005251896518187,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004883062843107207,
      "loss": 5.095,
      "mean_token_accuracy": 0.18799978792667388,
      "num_tokens": 24538593.0,
      "step": 14145
    },
    {
      "entropy": 5.661396884918213,
      "epoch": 1.1009142190235364,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882974639494383,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1773642987012863,
      "num_tokens": 24547158.0,
      "step": 14150
    },
    {
      "entropy": 5.539459991455078,
      "epoch": 1.1013032483952538,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004882886403517065,
      "loss": 5.1787,
      "mean_token_accuracy": 0.18391313701868056,
      "num_tokens": 24556266.0,
      "step": 14155
    },
    {
      "entropy": 5.6034904479980465,
      "epoch": 1.1016922777669713,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004882798135176589,
      "loss": 5.2333,
      "mean_token_accuracy": 0.18067197799682616,
      "num_tokens": 24565122.0,
      "step": 14160
    },
    {
      "entropy": 5.519475889205933,
      "epoch": 1.102081307138689,
      "grad_norm": 1.125,
      "learning_rate": 0.00048827098344742973,
      "loss": 5.0415,
      "mean_token_accuracy": 0.20198304504156112,
      "num_tokens": 24573944.0,
      "step": 14165
    },
    {
      "entropy": 5.609998083114624,
      "epoch": 1.1024703365104065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004882621501411528,
      "loss": 5.2464,
      "mean_token_accuracy": 0.18438272774219513,
      "num_tokens": 24582233.0,
      "step": 14170
    },
    {
      "entropy": 5.586512565612793,
      "epoch": 1.1028593658821242,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004882533135989622,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1918976992368698,
      "num_tokens": 24590855.0,
      "step": 14175
    },
    {
      "entropy": 5.6505523204803465,
      "epoch": 1.1032483952538417,
      "grad_norm": 1.09375,
      "learning_rate": 0.000488244473820992,
      "loss": 5.1648,
      "mean_token_accuracy": 0.17946069985628127,
      "num_tokens": 24599877.0,
      "step": 14180
    },
    {
      "entropy": 5.493092489242554,
      "epoch": 1.1036374246255591,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048823563080737634,
      "loss": 5.034,
      "mean_token_accuracy": 0.19380609542131425,
      "num_tokens": 24608450.0,
      "step": 14185
    },
    {
      "entropy": 5.4829517841339115,
      "epoch": 1.1040264539972768,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004882267845582493,
      "loss": 5.1143,
      "mean_token_accuracy": 0.1964149922132492,
      "num_tokens": 24618018.0,
      "step": 14190
    },
    {
      "entropy": 5.566144037246704,
      "epoch": 1.1044154833689943,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048821793507374526,
      "loss": 5.1024,
      "mean_token_accuracy": 0.1930859535932541,
      "num_tokens": 24626388.0,
      "step": 14195
    },
    {
      "entropy": 5.665661287307739,
      "epoch": 1.104804512740712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004882090823539984,
      "loss": 5.2258,
      "mean_token_accuracy": 0.18173695206642151,
      "num_tokens": 24635284.0,
      "step": 14200
    },
    {
      "entropy": 5.594166994094849,
      "epoch": 1.1051935421124295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004882002263991431,
      "loss": 5.0714,
      "mean_token_accuracy": 0.19396075010299682,
      "num_tokens": 24644297.0,
      "step": 14205
    },
    {
      "entropy": 5.45321249961853,
      "epoch": 1.105582571484147,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048819136720931364,
      "loss": 4.9489,
      "mean_token_accuracy": 0.19674341082572938,
      "num_tokens": 24652965.0,
      "step": 14210
    },
    {
      "entropy": 5.432612133026123,
      "epoch": 1.1059716008558647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048818250478464457,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2019434839487076,
      "num_tokens": 24661034.0,
      "step": 14215
    },
    {
      "entropy": 5.529057788848877,
      "epoch": 1.1063606302275821,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004881736391252703,
      "loss": 5.0988,
      "mean_token_accuracy": 0.19472705721855163,
      "num_tokens": 24669831.0,
      "step": 14220
    },
    {
      "entropy": 5.619505929946899,
      "epoch": 1.1067496595992998,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004881647702313253,
      "loss": 5.1584,
      "mean_token_accuracy": 0.1800076976418495,
      "num_tokens": 24678491.0,
      "step": 14225
    },
    {
      "entropy": 5.5571812152862545,
      "epoch": 1.1071386889710173,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048815589810294427,
      "loss": 5.1096,
      "mean_token_accuracy": 0.19239202737808228,
      "num_tokens": 24687308.0,
      "step": 14230
    },
    {
      "entropy": 5.503558158874512,
      "epoch": 1.1075277183427348,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048814702274026174,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19611527621746064,
      "num_tokens": 24696080.0,
      "step": 14235
    },
    {
      "entropy": 5.504922008514404,
      "epoch": 1.1079167477144525,
      "grad_norm": 1.125,
      "learning_rate": 0.0004881381441434124,
      "loss": 5.1569,
      "mean_token_accuracy": 0.187395840883255,
      "num_tokens": 24704561.0,
      "step": 14240
    },
    {
      "entropy": 5.60775055885315,
      "epoch": 1.10830577708617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000488129262312531,
      "loss": 5.1606,
      "mean_token_accuracy": 0.1901158019900322,
      "num_tokens": 24713403.0,
      "step": 14245
    },
    {
      "entropy": 5.607804679870606,
      "epoch": 1.1086948064578876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048812037724775217,
      "loss": 5.1619,
      "mean_token_accuracy": 0.1822177827358246,
      "num_tokens": 24722031.0,
      "step": 14250
    },
    {
      "entropy": 5.5430464267730715,
      "epoch": 1.1090838358296051,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004881114889492109,
      "loss": 5.1023,
      "mean_token_accuracy": 0.18657247126102447,
      "num_tokens": 24731176.0,
      "step": 14255
    },
    {
      "entropy": 5.581735849380493,
      "epoch": 1.1094728652013226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048810259741704197,
      "loss": 5.1437,
      "mean_token_accuracy": 0.18529002815485002,
      "num_tokens": 24740037.0,
      "step": 14260
    },
    {
      "entropy": 5.623476505279541,
      "epoch": 1.1098618945730403,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004880937026513803,
      "loss": 5.2081,
      "mean_token_accuracy": 0.18586902767419816,
      "num_tokens": 24748778.0,
      "step": 14265
    },
    {
      "entropy": 5.623208045959473,
      "epoch": 1.1102509239447578,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004880848046523609,
      "loss": 5.051,
      "mean_token_accuracy": 0.19547225087881087,
      "num_tokens": 24757226.0,
      "step": 14270
    },
    {
      "entropy": 5.509424161911011,
      "epoch": 1.1106399533164755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048807590342011875,
      "loss": 5.12,
      "mean_token_accuracy": 0.1936502203345299,
      "num_tokens": 24765734.0,
      "step": 14275
    },
    {
      "entropy": 5.503326320648194,
      "epoch": 1.111028982688193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048806699895478895,
      "loss": 5.0921,
      "mean_token_accuracy": 0.18585842847824097,
      "num_tokens": 24773773.0,
      "step": 14280
    },
    {
      "entropy": 5.546346044540405,
      "epoch": 1.1114180120599104,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004880580912565065,
      "loss": 5.1176,
      "mean_token_accuracy": 0.1898014023900032,
      "num_tokens": 24782563.0,
      "step": 14285
    },
    {
      "entropy": 5.589433479309082,
      "epoch": 1.1118070414316281,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004880491803254067,
      "loss": 5.1956,
      "mean_token_accuracy": 0.18674391359090806,
      "num_tokens": 24791492.0,
      "step": 14290
    },
    {
      "entropy": 5.52013521194458,
      "epoch": 1.1121960708033456,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004880402661616246,
      "loss": 5.0108,
      "mean_token_accuracy": 0.1970032900571823,
      "num_tokens": 24800229.0,
      "step": 14295
    },
    {
      "entropy": 5.542965316772461,
      "epoch": 1.1125851001750633,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004880313487652956,
      "loss": 5.0926,
      "mean_token_accuracy": 0.18638935089111328,
      "num_tokens": 24809454.0,
      "step": 14300
    },
    {
      "entropy": 5.507115221023559,
      "epoch": 1.1129741295467808,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048802242813655503,
      "loss": 5.0905,
      "mean_token_accuracy": 0.18912180811166762,
      "num_tokens": 24817956.0,
      "step": 14305
    },
    {
      "entropy": 5.497307443618775,
      "epoch": 1.1133631589184985,
      "grad_norm": 1.09375,
      "learning_rate": 0.000488013504275538,
      "loss": 5.0118,
      "mean_token_accuracy": 0.1968664661049843,
      "num_tokens": 24826288.0,
      "step": 14310
    },
    {
      "entropy": 5.605159091949463,
      "epoch": 1.113752188290216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004880045771823802,
      "loss": 5.1747,
      "mean_token_accuracy": 0.18741290271282196,
      "num_tokens": 24834720.0,
      "step": 14315
    },
    {
      "entropy": 5.566947889328003,
      "epoch": 1.1141412176619334,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048799564685721695,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1890547201037407,
      "num_tokens": 24843000.0,
      "step": 14320
    },
    {
      "entropy": 5.560391855239868,
      "epoch": 1.1145302470336511,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004879867133001837,
      "loss": 5.1165,
      "mean_token_accuracy": 0.1889778792858124,
      "num_tokens": 24851060.0,
      "step": 14325
    },
    {
      "entropy": 5.441255283355713,
      "epoch": 1.1149192764053686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004879777765114162,
      "loss": 5.0124,
      "mean_token_accuracy": 0.19799065887928008,
      "num_tokens": 24859913.0,
      "step": 14330
    },
    {
      "entropy": 5.599088478088379,
      "epoch": 1.115308305777086,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048796883649104996,
      "loss": 5.2372,
      "mean_token_accuracy": 0.18210665732622147,
      "num_tokens": 24869648.0,
      "step": 14335
    },
    {
      "entropy": 5.585079908370972,
      "epoch": 1.1156973351488038,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004879598932392206,
      "loss": 5.1306,
      "mean_token_accuracy": 0.19399481266736984,
      "num_tokens": 24878813.0,
      "step": 14340
    },
    {
      "entropy": 5.566272401809693,
      "epoch": 1.1160863645205212,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048795094675606364,
      "loss": 5.1294,
      "mean_token_accuracy": 0.1867375686764717,
      "num_tokens": 24887714.0,
      "step": 14345
    },
    {
      "entropy": 5.591864776611328,
      "epoch": 1.116475393892239,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004879419970417152,
      "loss": 5.108,
      "mean_token_accuracy": 0.18684172332286836,
      "num_tokens": 24895961.0,
      "step": 14350
    },
    {
      "entropy": 5.556972074508667,
      "epoch": 1.1168644232639564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048793304409631077,
      "loss": 5.1281,
      "mean_token_accuracy": 0.18404569625854492,
      "num_tokens": 24904873.0,
      "step": 14355
    },
    {
      "entropy": 5.488266086578369,
      "epoch": 1.117253452635674,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004879240879199863,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19572633802890776,
      "num_tokens": 24913486.0,
      "step": 14360
    },
    {
      "entropy": 5.544401025772094,
      "epoch": 1.1176424820073916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048791512851287786,
      "loss": 5.1335,
      "mean_token_accuracy": 0.18722374141216278,
      "num_tokens": 24921943.0,
      "step": 14365
    },
    {
      "entropy": 5.630269002914429,
      "epoch": 1.118031511379109,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048790616587512107,
      "loss": 5.2023,
      "mean_token_accuracy": 0.18350685834884645,
      "num_tokens": 24930081.0,
      "step": 14370
    },
    {
      "entropy": 5.628179740905762,
      "epoch": 1.1184205407508268,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004878972000068521,
      "loss": 5.122,
      "mean_token_accuracy": 0.19026199281215667,
      "num_tokens": 24938985.0,
      "step": 14375
    },
    {
      "entropy": 5.467494058609009,
      "epoch": 1.1188095701225442,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048788823090820707,
      "loss": 4.9755,
      "mean_token_accuracy": 0.19932196140289307,
      "num_tokens": 24947497.0,
      "step": 14380
    },
    {
      "entropy": 5.583019113540649,
      "epoch": 1.1191985994942617,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048787925857932194,
      "loss": 5.1835,
      "mean_token_accuracy": 0.18442970365285874,
      "num_tokens": 24956263.0,
      "step": 14385
    },
    {
      "entropy": 5.684103918075562,
      "epoch": 1.1195876288659794,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048787028302033287,
      "loss": 5.2767,
      "mean_token_accuracy": 0.18064552247524263,
      "num_tokens": 24965646.0,
      "step": 14390
    },
    {
      "entropy": 5.620402145385742,
      "epoch": 1.1199766582376969,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048786130423137606,
      "loss": 5.146,
      "mean_token_accuracy": 0.18637547492980958,
      "num_tokens": 24973318.0,
      "step": 14395
    },
    {
      "entropy": 5.581940317153931,
      "epoch": 1.1203656876094146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004878523222125879,
      "loss": 5.2118,
      "mean_token_accuracy": 0.18545670360326766,
      "num_tokens": 24982600.0,
      "step": 14400
    },
    {
      "entropy": 5.686072969436646,
      "epoch": 1.120754716981132,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004878433369641044,
      "loss": 5.2217,
      "mean_token_accuracy": 0.17989263832569122,
      "num_tokens": 24990386.0,
      "step": 14405
    },
    {
      "entropy": 5.475865697860717,
      "epoch": 1.1211437463528497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004878343484860621,
      "loss": 4.9889,
      "mean_token_accuracy": 0.19892134964466096,
      "num_tokens": 24999064.0,
      "step": 14410
    },
    {
      "entropy": 5.539119148254395,
      "epoch": 1.1215327757245672,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004878253567785972,
      "loss": 5.1716,
      "mean_token_accuracy": 0.1850363790988922,
      "num_tokens": 25008448.0,
      "step": 14415
    },
    {
      "entropy": 5.473070096969605,
      "epoch": 1.1219218050962847,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048781636184184644,
      "loss": 4.98,
      "mean_token_accuracy": 0.20469697266817094,
      "num_tokens": 25016722.0,
      "step": 14420
    },
    {
      "entropy": 5.494588088989258,
      "epoch": 1.1223108344680024,
      "grad_norm": 0.96875,
      "learning_rate": 0.000487807363675946,
      "loss": 5.0708,
      "mean_token_accuracy": 0.19545021802186965,
      "num_tokens": 25026263.0,
      "step": 14425
    },
    {
      "entropy": 5.573470497131348,
      "epoch": 1.1226998638397199,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004877983622810326,
      "loss": 5.1024,
      "mean_token_accuracy": 0.19335748255252838,
      "num_tokens": 25035202.0,
      "step": 14430
    },
    {
      "entropy": 5.610977840423584,
      "epoch": 1.1230888932114376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004877893576572427,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1878047689795494,
      "num_tokens": 25044309.0,
      "step": 14435
    },
    {
      "entropy": 5.592906665802002,
      "epoch": 1.123477922583155,
      "grad_norm": 1.0,
      "learning_rate": 0.000487780349804713,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19315307438373566,
      "num_tokens": 25053044.0,
      "step": 14440
    },
    {
      "entropy": 5.47516131401062,
      "epoch": 1.1238669519548725,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004877713387235802,
      "loss": 5.0755,
      "mean_token_accuracy": 0.19530935883522033,
      "num_tokens": 25061869.0,
      "step": 14445
    },
    {
      "entropy": 5.626252126693726,
      "epoch": 1.1242559813265902,
      "grad_norm": 1.15625,
      "learning_rate": 0.000487762324413981,
      "loss": 5.2573,
      "mean_token_accuracy": 0.18332788199186326,
      "num_tokens": 25070022.0,
      "step": 14450
    },
    {
      "entropy": 5.618163585662842,
      "epoch": 1.1246450106983077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048775330687605213,
      "loss": 5.1558,
      "mean_token_accuracy": 0.18399238884449004,
      "num_tokens": 25078604.0,
      "step": 14455
    },
    {
      "entropy": 5.528886461257935,
      "epoch": 1.1250340400700254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004877442861099305,
      "loss": 5.0085,
      "mean_token_accuracy": 0.19951817840337754,
      "num_tokens": 25087256.0,
      "step": 14460
    },
    {
      "entropy": 5.537620401382446,
      "epoch": 1.1254230694417429,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004877352621157529,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19643452912569045,
      "num_tokens": 25095593.0,
      "step": 14465
    },
    {
      "entropy": 5.574248600006103,
      "epoch": 1.1258120988134603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048772623489365635,
      "loss": 5.1394,
      "mean_token_accuracy": 0.18792365938425065,
      "num_tokens": 25104852.0,
      "step": 14470
    },
    {
      "entropy": 5.56501030921936,
      "epoch": 1.126201128185178,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004877172044437777,
      "loss": 5.1282,
      "mean_token_accuracy": 0.19328146129846574,
      "num_tokens": 25114453.0,
      "step": 14475
    },
    {
      "entropy": 5.572911882400513,
      "epoch": 1.1265901575568955,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048770817076625416,
      "loss": 5.123,
      "mean_token_accuracy": 0.19375506788492203,
      "num_tokens": 25123733.0,
      "step": 14480
    },
    {
      "entropy": 5.513044452667236,
      "epoch": 1.126979186928613,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048769913386122253,
      "loss": 5.044,
      "mean_token_accuracy": 0.19644900560379028,
      "num_tokens": 25132572.0,
      "step": 14485
    },
    {
      "entropy": 5.480722093582154,
      "epoch": 1.1273682163003307,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004876900937288202,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19857016950845718,
      "num_tokens": 25141494.0,
      "step": 14490
    },
    {
      "entropy": 5.502747488021851,
      "epoch": 1.1277572456720482,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048768105036918426,
      "loss": 5.0381,
      "mean_token_accuracy": 0.19095972329378127,
      "num_tokens": 25150186.0,
      "step": 14495
    },
    {
      "entropy": 5.561115694046021,
      "epoch": 1.1281462750437659,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048767200378245187,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19380782991647721,
      "num_tokens": 25159251.0,
      "step": 14500
    },
    {
      "entropy": 5.5895641326904295,
      "epoch": 1.1285353044154833,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048766295396876025,
      "loss": 5.1119,
      "mean_token_accuracy": 0.19401613026857376,
      "num_tokens": 25167780.0,
      "step": 14505
    },
    {
      "entropy": 5.521968936920166,
      "epoch": 1.128924333787201,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048765390092824683,
      "loss": 5.0392,
      "mean_token_accuracy": 0.19535757154226302,
      "num_tokens": 25175358.0,
      "step": 14510
    },
    {
      "entropy": 5.479304933547974,
      "epoch": 1.1293133631589185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000487644844661049,
      "loss": 5.0807,
      "mean_token_accuracy": 0.19168751388788224,
      "num_tokens": 25184247.0,
      "step": 14515
    },
    {
      "entropy": 5.478329133987427,
      "epoch": 1.129702392530636,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000487635785167304,
      "loss": 5.0144,
      "mean_token_accuracy": 0.19755800664424897,
      "num_tokens": 25192311.0,
      "step": 14520
    },
    {
      "entropy": 5.501252126693726,
      "epoch": 1.1300914219023537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048762672244714953,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20108913332223893,
      "num_tokens": 25200861.0,
      "step": 14525
    },
    {
      "entropy": 5.646828079223633,
      "epoch": 1.1304804512740712,
      "grad_norm": 1.125,
      "learning_rate": 0.0004876176565007229,
      "loss": 5.1895,
      "mean_token_accuracy": 0.1884510710835457,
      "num_tokens": 25208938.0,
      "step": 14530
    },
    {
      "entropy": 5.583438682556152,
      "epoch": 1.1308694806457888,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048760858732816174,
      "loss": 5.0856,
      "mean_token_accuracy": 0.19390414506196976,
      "num_tokens": 25216997.0,
      "step": 14535
    },
    {
      "entropy": 5.525766134262085,
      "epoch": 1.1312585100175063,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004875995149296037,
      "loss": 5.0067,
      "mean_token_accuracy": 0.19762847274541856,
      "num_tokens": 25225205.0,
      "step": 14540
    },
    {
      "entropy": 5.547775411605835,
      "epoch": 1.1316475393892238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004875904393051864,
      "loss": 5.0617,
      "mean_token_accuracy": 0.19728816151618958,
      "num_tokens": 25233807.0,
      "step": 14545
    },
    {
      "entropy": 5.530863666534424,
      "epoch": 1.1320365687609415,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004875813604550476,
      "loss": 5.0951,
      "mean_token_accuracy": 0.18459872603416444,
      "num_tokens": 25241715.0,
      "step": 14550
    },
    {
      "entropy": 5.6309418201446535,
      "epoch": 1.132425598132659,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048757227837932494,
      "loss": 5.2023,
      "mean_token_accuracy": 0.18823618143796922,
      "num_tokens": 25250418.0,
      "step": 14555
    },
    {
      "entropy": 5.6056373596191404,
      "epoch": 1.1328146275043767,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004875631930781563,
      "loss": 5.0986,
      "mean_token_accuracy": 0.18865856379270554,
      "num_tokens": 25259164.0,
      "step": 14560
    },
    {
      "entropy": 5.57382345199585,
      "epoch": 1.1332036568760941,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048755410455167967,
      "loss": 5.1404,
      "mean_token_accuracy": 0.19245496094226838,
      "num_tokens": 25267902.0,
      "step": 14565
    },
    {
      "entropy": 5.547601652145386,
      "epoch": 1.1335926862478116,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004875450128000326,
      "loss": 5.1069,
      "mean_token_accuracy": 0.1847388595342636,
      "num_tokens": 25275991.0,
      "step": 14570
    },
    {
      "entropy": 5.572788095474243,
      "epoch": 1.1339817156195293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004875359178233534,
      "loss": 5.1006,
      "mean_token_accuracy": 0.190210223197937,
      "num_tokens": 25284454.0,
      "step": 14575
    },
    {
      "entropy": 5.656353950500488,
      "epoch": 1.1343707449912468,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048752681962177985,
      "loss": 5.1981,
      "mean_token_accuracy": 0.19184107184410096,
      "num_tokens": 25292867.0,
      "step": 14580
    },
    {
      "entropy": 5.56125955581665,
      "epoch": 1.1347597743629645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048751771819545013,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1956092968583107,
      "num_tokens": 25301486.0,
      "step": 14585
    },
    {
      "entropy": 5.519770622253418,
      "epoch": 1.135148803734682,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048750861354450236,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1880646139383316,
      "num_tokens": 25310800.0,
      "step": 14590
    },
    {
      "entropy": 5.585312604904175,
      "epoch": 1.1355378331063994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004874995056690746,
      "loss": 5.1422,
      "mean_token_accuracy": 0.1884983852505684,
      "num_tokens": 25319063.0,
      "step": 14595
    },
    {
      "entropy": 5.685469341278076,
      "epoch": 1.1359268624781171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000487490394569305,
      "loss": 5.1899,
      "mean_token_accuracy": 0.18964117765426636,
      "num_tokens": 25327817.0,
      "step": 14600
    },
    {
      "entropy": 5.5358765602111815,
      "epoch": 1.1363158918498346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004874812802453319,
      "loss": 5.0377,
      "mean_token_accuracy": 0.18866246789693833,
      "num_tokens": 25335753.0,
      "step": 14605
    },
    {
      "entropy": 5.580042123794556,
      "epoch": 1.1367049212215523,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004874721626972936,
      "loss": 5.2019,
      "mean_token_accuracy": 0.18467870503664016,
      "num_tokens": 25344514.0,
      "step": 14610
    },
    {
      "entropy": 5.5758720397949215,
      "epoch": 1.1370939505932698,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004874630419253284,
      "loss": 5.0993,
      "mean_token_accuracy": 0.19326353073120117,
      "num_tokens": 25353003.0,
      "step": 14615
    },
    {
      "entropy": 5.547150897979736,
      "epoch": 1.1374829799649873,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048745391792957474,
      "loss": 5.1029,
      "mean_token_accuracy": 0.19042810797691345,
      "num_tokens": 25361599.0,
      "step": 14620
    },
    {
      "entropy": 5.58586368560791,
      "epoch": 1.137872009336705,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048744479071017097,
      "loss": 5.1358,
      "mean_token_accuracy": 0.1821556344628334,
      "num_tokens": 25370566.0,
      "step": 14625
    },
    {
      "entropy": 5.591298055648804,
      "epoch": 1.1382610387084224,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004874356602672556,
      "loss": 5.1467,
      "mean_token_accuracy": 0.18944391012191772,
      "num_tokens": 25379589.0,
      "step": 14630
    },
    {
      "entropy": 5.531908702850342,
      "epoch": 1.1386500680801401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004874265266009673,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19846725314855576,
      "num_tokens": 25387481.0,
      "step": 14635
    },
    {
      "entropy": 5.5358294486999515,
      "epoch": 1.1390390974518576,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004874173897114445,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1863221362233162,
      "num_tokens": 25395551.0,
      "step": 14640
    },
    {
      "entropy": 5.60767822265625,
      "epoch": 1.1394281268235753,
      "grad_norm": 1.125,
      "learning_rate": 0.00048740824959882603,
      "loss": 5.1628,
      "mean_token_accuracy": 0.1805470734834671,
      "num_tokens": 25404303.0,
      "step": 14645
    },
    {
      "entropy": 5.606921291351318,
      "epoch": 1.1398171561952928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873991062632504,
      "loss": 5.1631,
      "mean_token_accuracy": 0.1861985757946968,
      "num_tokens": 25412749.0,
      "step": 14650
    },
    {
      "entropy": 5.554032278060913,
      "epoch": 1.1402061855670103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048738995970485635,
      "loss": 5.0898,
      "mean_token_accuracy": 0.18690863996744156,
      "num_tokens": 25421075.0,
      "step": 14655
    },
    {
      "entropy": 5.588136482238769,
      "epoch": 1.140595214938728,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004873808099237827,
      "loss": 5.2237,
      "mean_token_accuracy": 0.1858045518398285,
      "num_tokens": 25429750.0,
      "step": 14660
    },
    {
      "entropy": 5.629778003692627,
      "epoch": 1.1409842443104454,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004873716569201684,
      "loss": 5.1162,
      "mean_token_accuracy": 0.1910490334033966,
      "num_tokens": 25437686.0,
      "step": 14665
    },
    {
      "entropy": 5.584583711624146,
      "epoch": 1.141373273682163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048736250069415213,
      "loss": 5.1131,
      "mean_token_accuracy": 0.19080674946308135,
      "num_tokens": 25446050.0,
      "step": 14670
    },
    {
      "entropy": 5.53749270439148,
      "epoch": 1.1417623030538806,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004873533412458729,
      "loss": 5.1035,
      "mean_token_accuracy": 0.19770489484071732,
      "num_tokens": 25455134.0,
      "step": 14675
    },
    {
      "entropy": 5.516529178619384,
      "epoch": 1.142151332425598,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048734417857546986,
      "loss": 5.0023,
      "mean_token_accuracy": 0.19790069460868837,
      "num_tokens": 25463258.0,
      "step": 14680
    },
    {
      "entropy": 5.582714796066284,
      "epoch": 1.1425403617973158,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004873350126830818,
      "loss": 5.0783,
      "mean_token_accuracy": 0.1942027196288109,
      "num_tokens": 25472085.0,
      "step": 14685
    },
    {
      "entropy": 5.582102918624878,
      "epoch": 1.1429293911690332,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004873258435688479,
      "loss": 5.088,
      "mean_token_accuracy": 0.1934540554881096,
      "num_tokens": 25481233.0,
      "step": 14690
    },
    {
      "entropy": 5.540214538574219,
      "epoch": 1.143318420540751,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004873166712329073,
      "loss": 5.0614,
      "mean_token_accuracy": 0.20070905983448029,
      "num_tokens": 25490354.0,
      "step": 14695
    },
    {
      "entropy": 5.540302419662476,
      "epoch": 1.1437074499124684,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048730749567539914,
      "loss": 5.1264,
      "mean_token_accuracy": 0.18700244128704072,
      "num_tokens": 25498467.0,
      "step": 14700
    },
    {
      "entropy": 5.585759973526001,
      "epoch": 1.144096479284186,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048729831689646257,
      "loss": 5.1373,
      "mean_token_accuracy": 0.19088697880506517,
      "num_tokens": 25507543.0,
      "step": 14705
    },
    {
      "entropy": 5.545436191558838,
      "epoch": 1.1444855086559036,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048728913489623705,
      "loss": 5.0279,
      "mean_token_accuracy": 0.1951140433549881,
      "num_tokens": 25516070.0,
      "step": 14710
    },
    {
      "entropy": 5.572587871551514,
      "epoch": 1.144874538027621,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004872799496748618,
      "loss": 5.1187,
      "mean_token_accuracy": 0.19245034754276275,
      "num_tokens": 25525216.0,
      "step": 14715
    },
    {
      "entropy": 5.581948900222779,
      "epoch": 1.1452635673993385,
      "grad_norm": 1.125,
      "learning_rate": 0.0004872707612324761,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1885973185300827,
      "num_tokens": 25533822.0,
      "step": 14720
    },
    {
      "entropy": 5.625873756408692,
      "epoch": 1.1456525967710562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004872615695692196,
      "loss": 5.1972,
      "mean_token_accuracy": 0.18007602840662001,
      "num_tokens": 25541580.0,
      "step": 14725
    },
    {
      "entropy": 5.494611978530884,
      "epoch": 1.1460416261427737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004872523746852315,
      "loss": 4.9929,
      "mean_token_accuracy": 0.19767958670854568,
      "num_tokens": 25549860.0,
      "step": 14730
    },
    {
      "entropy": 5.529557800292968,
      "epoch": 1.1464306555144914,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048724317658065146,
      "loss": 5.1403,
      "mean_token_accuracy": 0.18535635024309158,
      "num_tokens": 25559060.0,
      "step": 14735
    },
    {
      "entropy": 5.6726381301879885,
      "epoch": 1.146819684886209,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048723397525561916,
      "loss": 5.1968,
      "mean_token_accuracy": 0.18191540390253066,
      "num_tokens": 25567769.0,
      "step": 14740
    },
    {
      "entropy": 5.612449979782104,
      "epoch": 1.1472087142579266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048722477071027394,
      "loss": 5.1839,
      "mean_token_accuracy": 0.19144417047500611,
      "num_tokens": 25577443.0,
      "step": 14745
    },
    {
      "entropy": 5.551360130310059,
      "epoch": 1.147597743629644,
      "grad_norm": 1.0,
      "learning_rate": 0.00048721556294475574,
      "loss": 5.0977,
      "mean_token_accuracy": 0.1976507842540741,
      "num_tokens": 25586283.0,
      "step": 14750
    },
    {
      "entropy": 5.546954202651977,
      "epoch": 1.1479867730013615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048720635195920403,
      "loss": 5.1911,
      "mean_token_accuracy": 0.18866911381483079,
      "num_tokens": 25594881.0,
      "step": 14755
    },
    {
      "entropy": 5.53315372467041,
      "epoch": 1.1483758023730792,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004871971377537588,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19668822288513182,
      "num_tokens": 25603738.0,
      "step": 14760
    },
    {
      "entropy": 5.5818788528442385,
      "epoch": 1.1487648317447967,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004871879203285597,
      "loss": 5.1247,
      "mean_token_accuracy": 0.18973454385995864,
      "num_tokens": 25611780.0,
      "step": 14765
    },
    {
      "entropy": 5.591288900375366,
      "epoch": 1.1491538611165142,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048717869968374655,
      "loss": 5.1998,
      "mean_token_accuracy": 0.18770916312932967,
      "num_tokens": 25620489.0,
      "step": 14770
    },
    {
      "entropy": 5.522339105606079,
      "epoch": 1.1495428904882319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048716947581945954,
      "loss": 5.1378,
      "mean_token_accuracy": 0.19379799365997313,
      "num_tokens": 25629113.0,
      "step": 14775
    },
    {
      "entropy": 5.504387331008911,
      "epoch": 1.1499319198599494,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004871602487358383,
      "loss": 5.0848,
      "mean_token_accuracy": 0.18723666071891784,
      "num_tokens": 25637313.0,
      "step": 14780
    },
    {
      "entropy": 5.5924458503723145,
      "epoch": 1.150320949231667,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000487151018433023,
      "loss": 5.1912,
      "mean_token_accuracy": 0.1896853744983673,
      "num_tokens": 25646522.0,
      "step": 14785
    },
    {
      "entropy": 5.579239845275879,
      "epoch": 1.1507099786033845,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048714178491115367,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1924116864800453,
      "num_tokens": 25656058.0,
      "step": 14790
    },
    {
      "entropy": 5.652826261520386,
      "epoch": 1.1510990079751022,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048713254817037046,
      "loss": 5.1691,
      "mean_token_accuracy": 0.18834391981363297,
      "num_tokens": 25665072.0,
      "step": 14795
    },
    {
      "entropy": 5.621101760864258,
      "epoch": 1.1514880373468197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004871233082108135,
      "loss": 5.2249,
      "mean_token_accuracy": 0.18281933665275574,
      "num_tokens": 25673877.0,
      "step": 14800
    },
    {
      "entropy": 5.5880053520202635,
      "epoch": 1.1518770667185372,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871140650326228,
      "loss": 5.0836,
      "mean_token_accuracy": 0.18532530218362808,
      "num_tokens": 25683238.0,
      "step": 14805
    },
    {
      "entropy": 5.633425617218018,
      "epoch": 1.1522660960902549,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004871048186359389,
      "loss": 5.1785,
      "mean_token_accuracy": 0.18164470940828323,
      "num_tokens": 25692654.0,
      "step": 14810
    },
    {
      "entropy": 5.644726848602295,
      "epoch": 1.1526551254619724,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000487095569020902,
      "loss": 5.1583,
      "mean_token_accuracy": 0.18963952958583832,
      "num_tokens": 25701681.0,
      "step": 14815
    },
    {
      "entropy": 5.600336360931396,
      "epoch": 1.1530441548336898,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004870863161876524,
      "loss": 5.0678,
      "mean_token_accuracy": 0.19662757962942123,
      "num_tokens": 25709848.0,
      "step": 14820
    },
    {
      "entropy": 5.685653257369995,
      "epoch": 1.1534331842054075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004870770601363305,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1888419806957245,
      "num_tokens": 25717993.0,
      "step": 14825
    },
    {
      "entropy": 5.643045902252197,
      "epoch": 1.153822213577125,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048706780086707675,
      "loss": 5.2043,
      "mean_token_accuracy": 0.18233225792646407,
      "num_tokens": 25726379.0,
      "step": 14830
    },
    {
      "entropy": 5.542797231674195,
      "epoch": 1.1542112429488427,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004870585383800317,
      "loss": 5.1153,
      "mean_token_accuracy": 0.19497920125722884,
      "num_tokens": 25734968.0,
      "step": 14835
    },
    {
      "entropy": 5.588819265365601,
      "epoch": 1.1546002723205602,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048704927267533585,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19217586368322373,
      "num_tokens": 25743315.0,
      "step": 14840
    },
    {
      "entropy": 5.615221214294434,
      "epoch": 1.1549893016922779,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048704000375312975,
      "loss": 5.0758,
      "mean_token_accuracy": 0.18831450045108794,
      "num_tokens": 25752396.0,
      "step": 14845
    },
    {
      "entropy": 5.620538854598999,
      "epoch": 1.1553783310639953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048703073161355414,
      "loss": 5.1859,
      "mean_token_accuracy": 0.18929504603147507,
      "num_tokens": 25761142.0,
      "step": 14850
    },
    {
      "entropy": 5.511011743545533,
      "epoch": 1.1557673604357128,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004870214562567497,
      "loss": 5.0612,
      "mean_token_accuracy": 0.18714468330144882,
      "num_tokens": 25769514.0,
      "step": 14855
    },
    {
      "entropy": 5.543885707855225,
      "epoch": 1.1561563898074305,
      "grad_norm": 1.0625,
      "learning_rate": 0.000487012177682857,
      "loss": 5.0858,
      "mean_token_accuracy": 0.1892496019601822,
      "num_tokens": 25777408.0,
      "step": 14860
    },
    {
      "entropy": 5.5232421875,
      "epoch": 1.156545419179148,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004870028958920171,
      "loss": 5.052,
      "mean_token_accuracy": 0.19667184501886367,
      "num_tokens": 25785646.0,
      "step": 14865
    },
    {
      "entropy": 5.619859027862549,
      "epoch": 1.1569344485508655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004869936108843706,
      "loss": 5.1366,
      "mean_token_accuracy": 0.1919676572084427,
      "num_tokens": 25794007.0,
      "step": 14870
    },
    {
      "entropy": 5.595646858215332,
      "epoch": 1.1573234779225832,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048698432266005854,
      "loss": 5.0556,
      "mean_token_accuracy": 0.18877036571502687,
      "num_tokens": 25802765.0,
      "step": 14875
    },
    {
      "entropy": 5.626621341705322,
      "epoch": 1.1577125072943006,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004869750312192217,
      "loss": 5.2022,
      "mean_token_accuracy": 0.183294877409935,
      "num_tokens": 25811214.0,
      "step": 14880
    },
    {
      "entropy": 5.63381061553955,
      "epoch": 1.1581015366660183,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048696573656200123,
      "loss": 5.1907,
      "mean_token_accuracy": 0.18511330634355544,
      "num_tokens": 25820350.0,
      "step": 14885
    },
    {
      "entropy": 5.5866187572479244,
      "epoch": 1.1584905660377358,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048695643868853806,
      "loss": 5.107,
      "mean_token_accuracy": 0.19570998549461366,
      "num_tokens": 25828088.0,
      "step": 14890
    },
    {
      "entropy": 5.52763409614563,
      "epoch": 1.1588795954094535,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004869471375989733,
      "loss": 4.9949,
      "mean_token_accuracy": 0.19951591044664382,
      "num_tokens": 25835909.0,
      "step": 14895
    },
    {
      "entropy": 5.46763916015625,
      "epoch": 1.159268624781171,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004869378332934481,
      "loss": 5.0666,
      "mean_token_accuracy": 0.19120960384607316,
      "num_tokens": 25844360.0,
      "step": 14900
    },
    {
      "entropy": 5.631935024261475,
      "epoch": 1.1596576541528885,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048692852577210356,
      "loss": 5.1763,
      "mean_token_accuracy": 0.17828044295310974,
      "num_tokens": 25853877.0,
      "step": 14905
    },
    {
      "entropy": 5.57702145576477,
      "epoch": 1.1600466835246062,
      "grad_norm": 1.125,
      "learning_rate": 0.000486919215035081,
      "loss": 5.0682,
      "mean_token_accuracy": 0.19629686325788498,
      "num_tokens": 25862200.0,
      "step": 14910
    },
    {
      "entropy": 5.541756963729858,
      "epoch": 1.1604357128963236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004869099010825217,
      "loss": 5.0666,
      "mean_token_accuracy": 0.1891226202249527,
      "num_tokens": 25871112.0,
      "step": 14915
    },
    {
      "entropy": 5.5627213478088375,
      "epoch": 1.1608247422680413,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048690058391456684,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1875444993376732,
      "num_tokens": 25880135.0,
      "step": 14920
    },
    {
      "entropy": 5.570654582977295,
      "epoch": 1.1612137716397588,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004868912635313579,
      "loss": 5.1265,
      "mean_token_accuracy": 0.19106179773807525,
      "num_tokens": 25888609.0,
      "step": 14925
    },
    {
      "entropy": 5.569001865386963,
      "epoch": 1.1616028010114763,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004868819399330364,
      "loss": 5.1466,
      "mean_token_accuracy": 0.19086212962865828,
      "num_tokens": 25897254.0,
      "step": 14930
    },
    {
      "entropy": 5.582931280136108,
      "epoch": 1.161991830383194,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004868726131197436,
      "loss": 5.079,
      "mean_token_accuracy": 0.19257187843322754,
      "num_tokens": 25905705.0,
      "step": 14935
    },
    {
      "entropy": 5.59455361366272,
      "epoch": 1.1623808597549115,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004868632830916212,
      "loss": 5.1303,
      "mean_token_accuracy": 0.19274655878543853,
      "num_tokens": 25914066.0,
      "step": 14940
    },
    {
      "entropy": 5.5426572322845455,
      "epoch": 1.1627698891266292,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004868539498488106,
      "loss": 5.1312,
      "mean_token_accuracy": 0.18573394864797593,
      "num_tokens": 25922949.0,
      "step": 14945
    },
    {
      "entropy": 5.4484938621521,
      "epoch": 1.1631589184983466,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004868446133914536,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2006049409508705,
      "num_tokens": 25931465.0,
      "step": 14950
    },
    {
      "entropy": 5.567468690872192,
      "epoch": 1.163547947870064,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048683527371969185,
      "loss": 5.158,
      "mean_token_accuracy": 0.18369656354188918,
      "num_tokens": 25941014.0,
      "step": 14955
    },
    {
      "entropy": 5.527856636047363,
      "epoch": 1.1639369772417818,
      "grad_norm": 1.125,
      "learning_rate": 0.00048682593083366685,
      "loss": 5.0154,
      "mean_token_accuracy": 0.1881873354315758,
      "num_tokens": 25950275.0,
      "step": 14960
    },
    {
      "entropy": 5.5191387176513675,
      "epoch": 1.1643260066134993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048681658473352055,
      "loss": 5.1611,
      "mean_token_accuracy": 0.1887619227170944,
      "num_tokens": 25959138.0,
      "step": 14965
    },
    {
      "entropy": 5.587380313873291,
      "epoch": 1.164715035985217,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004868072354193947,
      "loss": 5.0286,
      "mean_token_accuracy": 0.1940028801560402,
      "num_tokens": 25967221.0,
      "step": 14970
    },
    {
      "entropy": 5.570781898498535,
      "epoch": 1.1651040653569344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004867978828914312,
      "loss": 5.131,
      "mean_token_accuracy": 0.19069798290729523,
      "num_tokens": 25976194.0,
      "step": 14975
    },
    {
      "entropy": 5.545144701004029,
      "epoch": 1.165493094728652,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004867885271497719,
      "loss": 5.1,
      "mean_token_accuracy": 0.18754445910453796,
      "num_tokens": 25985493.0,
      "step": 14980
    },
    {
      "entropy": 5.608636999130249,
      "epoch": 1.1658821241003696,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004867791681945588,
      "loss": 5.0766,
      "mean_token_accuracy": 0.1822480782866478,
      "num_tokens": 25994213.0,
      "step": 14985
    },
    {
      "entropy": 5.518993806838989,
      "epoch": 1.166271153472087,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048676980602593395,
      "loss": 5.0443,
      "mean_token_accuracy": 0.19690360277891159,
      "num_tokens": 26003164.0,
      "step": 14990
    },
    {
      "entropy": 5.569867372512817,
      "epoch": 1.1666601828438048,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048676044064403936,
      "loss": 5.1393,
      "mean_token_accuracy": 0.18535830974578857,
      "num_tokens": 26012049.0,
      "step": 14995
    },
    {
      "entropy": 5.684781074523926,
      "epoch": 1.1670492122155223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004867510720490171,
      "loss": 5.252,
      "mean_token_accuracy": 0.18107686638832093,
      "num_tokens": 26020466.0,
      "step": 15000
    },
    {
      "epoch": 1.1670492122155223,
      "eval_entropy": 5.416919413796976,
      "eval_loss": 5.200766086578369,
      "eval_mean_token_accuracy": 0.19595253605097962,
      "eval_num_tokens": 26020466.0,
      "eval_runtime": 21.865,
      "eval_samples_per_second": 1900.663,
      "eval_steps_per_second": 237.594,
      "step": 15000
    },
    {
      "entropy": 5.541977643966675,
      "epoch": 1.1674382415872397,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004867417002410094,
      "loss": 5.0382,
      "mean_token_accuracy": 0.19401247799396515,
      "num_tokens": 26028815.0,
      "step": 15005
    },
    {
      "entropy": 5.597784614562988,
      "epoch": 1.1678272709589574,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048673232522015845,
      "loss": 5.2426,
      "mean_token_accuracy": 0.1854727789759636,
      "num_tokens": 26037759.0,
      "step": 15010
    },
    {
      "entropy": 5.585027074813842,
      "epoch": 1.168216300330675,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004867229469866064,
      "loss": 5.1554,
      "mean_token_accuracy": 0.18069108873605727,
      "num_tokens": 26046349.0,
      "step": 15015
    },
    {
      "entropy": 5.624474143981933,
      "epoch": 1.1686053297023926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867135655404956,
      "loss": 5.1385,
      "mean_token_accuracy": 0.19369989186525344,
      "num_tokens": 26055265.0,
      "step": 15020
    },
    {
      "entropy": 5.6418193817138675,
      "epoch": 1.16899435907411,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048670418088196844,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19563569575548173,
      "num_tokens": 26064335.0,
      "step": 15025
    },
    {
      "entropy": 5.575631189346313,
      "epoch": 1.1693833884458276,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004866947930111674,
      "loss": 5.0522,
      "mean_token_accuracy": 0.19849299341440202,
      "num_tokens": 26073044.0,
      "step": 15030
    },
    {
      "entropy": 5.543219375610351,
      "epoch": 1.1697724178175453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048668540192823467,
      "loss": 5.0918,
      "mean_token_accuracy": 0.19380704909563065,
      "num_tokens": 26081361.0,
      "step": 15035
    },
    {
      "entropy": 5.545398902893067,
      "epoch": 1.1701614471892627,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000486676007633313,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19307537227869034,
      "num_tokens": 26089740.0,
      "step": 15040
    },
    {
      "entropy": 5.643700504302979,
      "epoch": 1.1705504765609804,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004866666101265448,
      "loss": 5.1704,
      "mean_token_accuracy": 0.19129014909267425,
      "num_tokens": 26098906.0,
      "step": 15045
    },
    {
      "entropy": 5.602594375610352,
      "epoch": 1.170939505932698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048665720940807274,
      "loss": 5.1633,
      "mean_token_accuracy": 0.18365934491157532,
      "num_tokens": 26107978.0,
      "step": 15050
    },
    {
      "entropy": 5.557853174209595,
      "epoch": 1.1713285353044154,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048664780547803935,
      "loss": 5.078,
      "mean_token_accuracy": 0.18788159042596816,
      "num_tokens": 26116756.0,
      "step": 15055
    },
    {
      "entropy": 5.657771921157837,
      "epoch": 1.171717564676133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004866383983365873,
      "loss": 5.151,
      "mean_token_accuracy": 0.18367486745119094,
      "num_tokens": 26125078.0,
      "step": 15060
    },
    {
      "entropy": 5.632147121429443,
      "epoch": 1.1721065940478506,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048662898798385956,
      "loss": 5.189,
      "mean_token_accuracy": 0.18557200729846954,
      "num_tokens": 26134149.0,
      "step": 15065
    },
    {
      "entropy": 5.566326522827149,
      "epoch": 1.1724956234195683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048661957441999875,
      "loss": 5.0571,
      "mean_token_accuracy": 0.1970219224691391,
      "num_tokens": 26142789.0,
      "step": 15070
    },
    {
      "entropy": 5.5901477336883545,
      "epoch": 1.1728846527912857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048661015764514765,
      "loss": 5.2055,
      "mean_token_accuracy": 0.18510168194770812,
      "num_tokens": 26151189.0,
      "step": 15075
    },
    {
      "entropy": 5.561592721939087,
      "epoch": 1.1732736821630034,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048660073765944927,
      "loss": 5.1164,
      "mean_token_accuracy": 0.19420742690563203,
      "num_tokens": 26159495.0,
      "step": 15080
    },
    {
      "entropy": 5.550236558914184,
      "epoch": 1.173662711534721,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004865913144630465,
      "loss": 5.0339,
      "mean_token_accuracy": 0.19319519847631456,
      "num_tokens": 26167804.0,
      "step": 15085
    },
    {
      "entropy": 5.582855653762818,
      "epoch": 1.1740517409064384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048658188805608227,
      "loss": 5.0855,
      "mean_token_accuracy": 0.191781710088253,
      "num_tokens": 26176240.0,
      "step": 15090
    },
    {
      "entropy": 5.569748640060425,
      "epoch": 1.174440770278156,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004865724584386998,
      "loss": 5.0924,
      "mean_token_accuracy": 0.1925620034337044,
      "num_tokens": 26184580.0,
      "step": 15095
    },
    {
      "entropy": 5.566873550415039,
      "epoch": 1.1748297996498736,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004865630256110418,
      "loss": 5.1334,
      "mean_token_accuracy": 0.1881118357181549,
      "num_tokens": 26193756.0,
      "step": 15100
    },
    {
      "entropy": 5.623624324798584,
      "epoch": 1.175218829021591,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048655358957325174,
      "loss": 5.1438,
      "mean_token_accuracy": 0.1903481274843216,
      "num_tokens": 26201818.0,
      "step": 15105
    },
    {
      "entropy": 5.5879504680633545,
      "epoch": 1.1756078583933087,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048654415032547277,
      "loss": 5.2254,
      "mean_token_accuracy": 0.17875734865665435,
      "num_tokens": 26211817.0,
      "step": 15110
    },
    {
      "entropy": 5.617209577560425,
      "epoch": 1.1759968877650262,
      "grad_norm": 1.125,
      "learning_rate": 0.00048653470786784794,
      "loss": 5.0793,
      "mean_token_accuracy": 0.18683170527219772,
      "num_tokens": 26220280.0,
      "step": 15115
    },
    {
      "entropy": 5.4996977806091305,
      "epoch": 1.176385917136744,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004865252622005206,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2011154055595398,
      "num_tokens": 26228719.0,
      "step": 15120
    },
    {
      "entropy": 5.627125597000122,
      "epoch": 1.1767749465084614,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048651581332363405,
      "loss": 5.135,
      "mean_token_accuracy": 0.1885628193616867,
      "num_tokens": 26237462.0,
      "step": 15125
    },
    {
      "entropy": 5.667350482940674,
      "epoch": 1.177163975880179,
      "grad_norm": 1.125,
      "learning_rate": 0.00048650636123733176,
      "loss": 5.223,
      "mean_token_accuracy": 0.1903729557991028,
      "num_tokens": 26245747.0,
      "step": 15130
    },
    {
      "entropy": 5.627001047134399,
      "epoch": 1.1775530052518965,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004864969059417571,
      "loss": 5.093,
      "mean_token_accuracy": 0.19148961901664735,
      "num_tokens": 26254977.0,
      "step": 15135
    },
    {
      "entropy": 5.582975196838379,
      "epoch": 1.177942034623614,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004864874474370534,
      "loss": 5.2269,
      "mean_token_accuracy": 0.18157656639814376,
      "num_tokens": 26264298.0,
      "step": 15140
    },
    {
      "entropy": 5.625980424880981,
      "epoch": 1.1783310639953317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048647798572336445,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18855575174093248,
      "num_tokens": 26273454.0,
      "step": 15145
    },
    {
      "entropy": 5.623959302902222,
      "epoch": 1.1787200933670492,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048646852080083355,
      "loss": 5.1112,
      "mean_token_accuracy": 0.19174898713827132,
      "num_tokens": 26282584.0,
      "step": 15150
    },
    {
      "entropy": 5.579500532150268,
      "epoch": 1.1791091227387667,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004864590526696045,
      "loss": 5.1496,
      "mean_token_accuracy": 0.18989068567752837,
      "num_tokens": 26291633.0,
      "step": 15155
    },
    {
      "entropy": 5.624048566818237,
      "epoch": 1.1794981521104844,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048644958132982094,
      "loss": 5.1811,
      "mean_token_accuracy": 0.18464895486831664,
      "num_tokens": 26300808.0,
      "step": 15160
    },
    {
      "entropy": 5.567554903030396,
      "epoch": 1.1798871814822018,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048644010678162644,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19849553108215331,
      "num_tokens": 26308976.0,
      "step": 15165
    },
    {
      "entropy": 5.610508966445923,
      "epoch": 1.1802762108539195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004864306290251649,
      "loss": 5.1291,
      "mean_token_accuracy": 0.18922855108976364,
      "num_tokens": 26316980.0,
      "step": 15170
    },
    {
      "entropy": 5.546866512298584,
      "epoch": 1.180665240225637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048642114806058014,
      "loss": 5.036,
      "mean_token_accuracy": 0.19938744753599166,
      "num_tokens": 26325723.0,
      "step": 15175
    },
    {
      "entropy": 5.5851257801055905,
      "epoch": 1.1810542695973547,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004864116638880159,
      "loss": 5.1442,
      "mean_token_accuracy": 0.18756299316883088,
      "num_tokens": 26334851.0,
      "step": 15180
    },
    {
      "entropy": 5.596612119674683,
      "epoch": 1.1814432989690722,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004864021765076162,
      "loss": 5.142,
      "mean_token_accuracy": 0.1861535996198654,
      "num_tokens": 26343696.0,
      "step": 15185
    },
    {
      "entropy": 5.649726152420044,
      "epoch": 1.1818323283407897,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048639268591952494,
      "loss": 5.2527,
      "mean_token_accuracy": 0.18580673933029174,
      "num_tokens": 26353054.0,
      "step": 15190
    },
    {
      "entropy": 5.586756563186645,
      "epoch": 1.1822213577125074,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048638319212388614,
      "loss": 5.0929,
      "mean_token_accuracy": 0.1926242157816887,
      "num_tokens": 26361484.0,
      "step": 15195
    },
    {
      "entropy": 5.624805498123169,
      "epoch": 1.1826103870842248,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004863736951208438,
      "loss": 5.2008,
      "mean_token_accuracy": 0.180966754257679,
      "num_tokens": 26370402.0,
      "step": 15200
    },
    {
      "entropy": 5.626142263412476,
      "epoch": 1.1829994164559423,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048636419491054217,
      "loss": 5.2292,
      "mean_token_accuracy": 0.17904138565063477,
      "num_tokens": 26378746.0,
      "step": 15205
    },
    {
      "entropy": 5.5672883033752445,
      "epoch": 1.18338844582766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004863546914931252,
      "loss": 5.0903,
      "mean_token_accuracy": 0.196580870449543,
      "num_tokens": 26386984.0,
      "step": 15210
    },
    {
      "entropy": 5.670325183868409,
      "epoch": 1.1837774751993775,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004863451848687373,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18338609784841536,
      "num_tokens": 26395712.0,
      "step": 15215
    },
    {
      "entropy": 5.562811756134034,
      "epoch": 1.1841665045710952,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048633567503752253,
      "loss": 5.1408,
      "mean_token_accuracy": 0.1894928827881813,
      "num_tokens": 26406421.0,
      "step": 15220
    },
    {
      "entropy": 5.619890022277832,
      "epoch": 1.1845555339428127,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004863261619996253,
      "loss": 5.1361,
      "mean_token_accuracy": 0.19117431789636613,
      "num_tokens": 26416032.0,
      "step": 15225
    },
    {
      "entropy": 5.589944505691529,
      "epoch": 1.1849445633145304,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004863166457551899,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19094704687595368,
      "num_tokens": 26423903.0,
      "step": 15230
    },
    {
      "entropy": 5.575456380844116,
      "epoch": 1.1853335926862478,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004863071263043608,
      "loss": 5.1634,
      "mean_token_accuracy": 0.18706756830215454,
      "num_tokens": 26432631.0,
      "step": 15235
    },
    {
      "entropy": 5.580493307113647,
      "epoch": 1.1857226220579653,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004862976036472823,
      "loss": 5.1617,
      "mean_token_accuracy": 0.18925195783376694,
      "num_tokens": 26441529.0,
      "step": 15240
    },
    {
      "entropy": 5.5786971092224125,
      "epoch": 1.186111651429683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048628807778409907,
      "loss": 5.0676,
      "mean_token_accuracy": 0.19940820336341858,
      "num_tokens": 26449928.0,
      "step": 15245
    },
    {
      "entropy": 5.555418348312378,
      "epoch": 1.1865006808014005,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004862785487149555,
      "loss": 5.0528,
      "mean_token_accuracy": 0.19879109114408494,
      "num_tokens": 26458065.0,
      "step": 15250
    },
    {
      "entropy": 5.558574581146241,
      "epoch": 1.186889710173118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004862690164399963,
      "loss": 5.1055,
      "mean_token_accuracy": 0.18898257315158845,
      "num_tokens": 26466690.0,
      "step": 15255
    },
    {
      "entropy": 5.542402553558349,
      "epoch": 1.1872787395448356,
      "grad_norm": 1.125,
      "learning_rate": 0.00048625948095936593,
      "loss": 5.1544,
      "mean_token_accuracy": 0.1839495614171028,
      "num_tokens": 26476133.0,
      "step": 15260
    },
    {
      "entropy": 5.616349458694458,
      "epoch": 1.1876677689165531,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048624994227320923,
      "loss": 5.1181,
      "mean_token_accuracy": 0.19102234542369842,
      "num_tokens": 26485151.0,
      "step": 15265
    },
    {
      "entropy": 5.600252103805542,
      "epoch": 1.1880567982882708,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048624040038167094,
      "loss": 5.1467,
      "mean_token_accuracy": 0.188801546394825,
      "num_tokens": 26494178.0,
      "step": 15270
    },
    {
      "entropy": 5.607261037826538,
      "epoch": 1.1884458276599883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048623085528489584,
      "loss": 5.1414,
      "mean_token_accuracy": 0.19747592508792877,
      "num_tokens": 26503233.0,
      "step": 15275
    },
    {
      "entropy": 5.559400415420532,
      "epoch": 1.188834857031706,
      "grad_norm": 1.0,
      "learning_rate": 0.00048622130698302863,
      "loss": 5.007,
      "mean_token_accuracy": 0.19970201849937438,
      "num_tokens": 26512127.0,
      "step": 15280
    },
    {
      "entropy": 5.562219429016113,
      "epoch": 1.1892238864034235,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048621175547621426,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20449354499578476,
      "num_tokens": 26520239.0,
      "step": 15285
    },
    {
      "entropy": 5.479774427413941,
      "epoch": 1.189612915775141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004862022007645978,
      "loss": 5.1252,
      "mean_token_accuracy": 0.18971802443265914,
      "num_tokens": 26528946.0,
      "step": 15290
    },
    {
      "entropy": 5.57308406829834,
      "epoch": 1.1900019451468586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048619264284832403,
      "loss": 5.1365,
      "mean_token_accuracy": 0.18855781704187394,
      "num_tokens": 26536825.0,
      "step": 15295
    },
    {
      "entropy": 5.609621667861939,
      "epoch": 1.1903909745185761,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048618308172753804,
      "loss": 5.0733,
      "mean_token_accuracy": 0.1971067577600479,
      "num_tokens": 26545208.0,
      "step": 15300
    },
    {
      "entropy": 5.571625137329102,
      "epoch": 1.1907800038902936,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004861735174023849,
      "loss": 5.0854,
      "mean_token_accuracy": 0.1926027461886406,
      "num_tokens": 26553768.0,
      "step": 15305
    },
    {
      "entropy": 5.619600677490235,
      "epoch": 1.1911690332620113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048616394987300985,
      "loss": 5.1437,
      "mean_token_accuracy": 0.1917687714099884,
      "num_tokens": 26561582.0,
      "step": 15310
    },
    {
      "entropy": 5.54694709777832,
      "epoch": 1.1915580626337288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004861543791395579,
      "loss": 5.0787,
      "mean_token_accuracy": 0.1894313797354698,
      "num_tokens": 26570632.0,
      "step": 15315
    },
    {
      "entropy": 5.5129759311676025,
      "epoch": 1.1919470920054465,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004861448052021743,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19717872142791748,
      "num_tokens": 26579507.0,
      "step": 15320
    },
    {
      "entropy": 5.565240955352783,
      "epoch": 1.192336121377164,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004861352280610044,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19538176506757737,
      "num_tokens": 26588765.0,
      "step": 15325
    },
    {
      "entropy": 5.630602788925171,
      "epoch": 1.1927251507488816,
      "grad_norm": 1.0,
      "learning_rate": 0.00048612564771619346,
      "loss": 5.1487,
      "mean_token_accuracy": 0.18356246501207352,
      "num_tokens": 26597529.0,
      "step": 15330
    },
    {
      "entropy": 5.52848801612854,
      "epoch": 1.193114180120599,
      "grad_norm": 1.0,
      "learning_rate": 0.00048611606416788686,
      "loss": 5.0823,
      "mean_token_accuracy": 0.19490896314382553,
      "num_tokens": 26605960.0,
      "step": 15335
    },
    {
      "entropy": 5.534823656082153,
      "epoch": 1.1935032094923166,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048610647741622996,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20490020662546157,
      "num_tokens": 26614435.0,
      "step": 15340
    },
    {
      "entropy": 5.557860374450684,
      "epoch": 1.1938922388640343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004860968874613683,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19326656311750412,
      "num_tokens": 26622873.0,
      "step": 15345
    },
    {
      "entropy": 5.544766139984131,
      "epoch": 1.1942812682357518,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004860872943034474,
      "loss": 5.056,
      "mean_token_accuracy": 0.197356016933918,
      "num_tokens": 26631529.0,
      "step": 15350
    },
    {
      "entropy": 5.604077911376953,
      "epoch": 1.1946702976074695,
      "grad_norm": 1.125,
      "learning_rate": 0.0004860776979426128,
      "loss": 5.1584,
      "mean_token_accuracy": 0.1809865191578865,
      "num_tokens": 26639466.0,
      "step": 15355
    },
    {
      "entropy": 5.496183681488037,
      "epoch": 1.195059326979187,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004860680983790101,
      "loss": 5.1081,
      "mean_token_accuracy": 0.1902945876121521,
      "num_tokens": 26648329.0,
      "step": 15360
    },
    {
      "entropy": 5.583552694320678,
      "epoch": 1.1954483563509044,
      "grad_norm": 1.0,
      "learning_rate": 0.0004860584956127849,
      "loss": 5.0315,
      "mean_token_accuracy": 0.19532059133052826,
      "num_tokens": 26656899.0,
      "step": 15365
    },
    {
      "entropy": 5.671739912033081,
      "epoch": 1.195837385722622,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048604888964408306,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1841220363974571,
      "num_tokens": 26666155.0,
      "step": 15370
    },
    {
      "entropy": 5.525110530853271,
      "epoch": 1.1962264150943396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004860392804730502,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1970916911959648,
      "num_tokens": 26675567.0,
      "step": 15375
    },
    {
      "entropy": 5.558688926696777,
      "epoch": 1.1966154444660573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048602966809983204,
      "loss": 5.0871,
      "mean_token_accuracy": 0.1909484878182411,
      "num_tokens": 26684133.0,
      "step": 15380
    },
    {
      "entropy": 5.580195140838623,
      "epoch": 1.1970044738377748,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004860200525245746,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18196916282176973,
      "num_tokens": 26692746.0,
      "step": 15385
    },
    {
      "entropy": 5.603961038589477,
      "epoch": 1.1973935032094922,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004860104337474239,
      "loss": 5.1085,
      "mean_token_accuracy": 0.19260500818490983,
      "num_tokens": 26701887.0,
      "step": 15390
    },
    {
      "entropy": 5.565649843215942,
      "epoch": 1.19778253258121,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048600081176852555,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19033753126859665,
      "num_tokens": 26710346.0,
      "step": 15395
    },
    {
      "entropy": 5.519503259658814,
      "epoch": 1.1981715619529274,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048599118658802575,
      "loss": 5.0129,
      "mean_token_accuracy": 0.19562781006097793,
      "num_tokens": 26718775.0,
      "step": 15400
    },
    {
      "entropy": 5.511463689804077,
      "epoch": 1.198560591324645,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004859815582060706,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19040385633707047,
      "num_tokens": 26726683.0,
      "step": 15405
    },
    {
      "entropy": 5.5324687480926515,
      "epoch": 1.1989496206963626,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004859719266228061,
      "loss": 5.0821,
      "mean_token_accuracy": 0.1944702908396721,
      "num_tokens": 26735450.0,
      "step": 15410
    },
    {
      "entropy": 5.580057811737061,
      "epoch": 1.19933865006808,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004859622918383784,
      "loss": 5.0386,
      "mean_token_accuracy": 0.19432027041912078,
      "num_tokens": 26744006.0,
      "step": 15415
    },
    {
      "entropy": 5.494427108764649,
      "epoch": 1.1997276794397977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859526538529337,
      "loss": 4.9841,
      "mean_token_accuracy": 0.20163743048906327,
      "num_tokens": 26751964.0,
      "step": 15420
    },
    {
      "entropy": 5.485170698165893,
      "epoch": 1.2001167088115152,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004859430126666182,
      "loss": 5.0434,
      "mean_token_accuracy": 0.1918250784277916,
      "num_tokens": 26760766.0,
      "step": 15425
    },
    {
      "entropy": 5.517698335647583,
      "epoch": 1.200505738183233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048593336827957824,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19140982776880264,
      "num_tokens": 26769663.0,
      "step": 15430
    },
    {
      "entropy": 5.641649341583252,
      "epoch": 1.2008947675549504,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048592372069196024,
      "loss": 5.2477,
      "mean_token_accuracy": 0.17693831473588945,
      "num_tokens": 26778693.0,
      "step": 15435
    },
    {
      "entropy": 5.5501556396484375,
      "epoch": 1.2012837969266679,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004859140699039104,
      "loss": 5.027,
      "mean_token_accuracy": 0.20334301888942719,
      "num_tokens": 26787368.0,
      "step": 15440
    },
    {
      "entropy": 5.499142360687256,
      "epoch": 1.2016728262983856,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048590441591557526,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20711124837398528,
      "num_tokens": 26795559.0,
      "step": 15445
    },
    {
      "entropy": 5.486826086044312,
      "epoch": 1.202061855670103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048589475872710134,
      "loss": 5.0051,
      "mean_token_accuracy": 0.19876922070980071,
      "num_tokens": 26803657.0,
      "step": 15450
    },
    {
      "entropy": 5.617013025283813,
      "epoch": 1.2024508850418207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004858850983386351,
      "loss": 5.1882,
      "mean_token_accuracy": 0.18423074781894683,
      "num_tokens": 26812214.0,
      "step": 15455
    },
    {
      "entropy": 5.594395303726197,
      "epoch": 1.2028399144135382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004858754347503231,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2038549467921257,
      "num_tokens": 26820591.0,
      "step": 15460
    },
    {
      "entropy": 5.533780431747436,
      "epoch": 1.203228943785256,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048586576796231216,
      "loss": 5.1258,
      "mean_token_accuracy": 0.18946319073438644,
      "num_tokens": 26829228.0,
      "step": 15465
    },
    {
      "entropy": 5.626135349273682,
      "epoch": 1.2036179731569734,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004858560979747487,
      "loss": 5.2631,
      "mean_token_accuracy": 0.1766286924481392,
      "num_tokens": 26838083.0,
      "step": 15470
    },
    {
      "entropy": 5.539951181411743,
      "epoch": 1.2040070025286909,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048584642478777967,
      "loss": 4.9582,
      "mean_token_accuracy": 0.20000711530447007,
      "num_tokens": 26846796.0,
      "step": 15475
    },
    {
      "entropy": 5.520544910430909,
      "epoch": 1.2043960319004086,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004858367484015517,
      "loss": 5.0771,
      "mean_token_accuracy": 0.1925956755876541,
      "num_tokens": 26854874.0,
      "step": 15480
    },
    {
      "entropy": 5.5402092933654785,
      "epoch": 1.204785061272126,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048582706881621166,
      "loss": 5.1074,
      "mean_token_accuracy": 0.1899310678243637,
      "num_tokens": 26863040.0,
      "step": 15485
    },
    {
      "entropy": 5.582753610610962,
      "epoch": 1.2051740906438435,
      "grad_norm": 1.125,
      "learning_rate": 0.0004858173860319063,
      "loss": 5.0991,
      "mean_token_accuracy": 0.18925801366567613,
      "num_tokens": 26872183.0,
      "step": 15490
    },
    {
      "entropy": 5.6179900646209715,
      "epoch": 1.2055631200155612,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048580770004878277,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18819399625062944,
      "num_tokens": 26880480.0,
      "step": 15495
    },
    {
      "entropy": 5.564848852157593,
      "epoch": 1.2059521493872787,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004857980108669879,
      "loss": 5.101,
      "mean_token_accuracy": 0.19431786239147186,
      "num_tokens": 26888855.0,
      "step": 15500
    },
    {
      "entropy": 5.553756904602051,
      "epoch": 1.2063411787589964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048578831848666875,
      "loss": 5.1103,
      "mean_token_accuracy": 0.18740586638450624,
      "num_tokens": 26897692.0,
      "step": 15505
    },
    {
      "entropy": 5.59758939743042,
      "epoch": 1.2067302081307139,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048577862290797235,
      "loss": 5.0584,
      "mean_token_accuracy": 0.19087831676006317,
      "num_tokens": 26906502.0,
      "step": 15510
    },
    {
      "entropy": 5.603816747665405,
      "epoch": 1.2071192375024316,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048576892413104586,
      "loss": 5.1126,
      "mean_token_accuracy": 0.1938114881515503,
      "num_tokens": 26915246.0,
      "step": 15515
    },
    {
      "entropy": 5.63941125869751,
      "epoch": 1.207508266874149,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004857592221560364,
      "loss": 5.2044,
      "mean_token_accuracy": 0.19181982278823853,
      "num_tokens": 26923901.0,
      "step": 15520
    },
    {
      "entropy": 5.613782501220703,
      "epoch": 1.2078972962458665,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004857495169830912,
      "loss": 5.2813,
      "mean_token_accuracy": 0.18695232421159744,
      "num_tokens": 26933246.0,
      "step": 15525
    },
    {
      "entropy": 5.584592533111572,
      "epoch": 1.2082863256175842,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004857398086123575,
      "loss": 5.1235,
      "mean_token_accuracy": 0.19786610454320908,
      "num_tokens": 26941938.0,
      "step": 15530
    },
    {
      "entropy": 5.575952005386353,
      "epoch": 1.2086753549893017,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004857300970439827,
      "loss": 5.0842,
      "mean_token_accuracy": 0.1901939332485199,
      "num_tokens": 26949967.0,
      "step": 15535
    },
    {
      "entropy": 5.594762659072876,
      "epoch": 1.2090643843610192,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048572038227811404,
      "loss": 5.1481,
      "mean_token_accuracy": 0.1893602952361107,
      "num_tokens": 26958743.0,
      "step": 15540
    },
    {
      "entropy": 5.664062976837158,
      "epoch": 1.2094534137327368,
      "grad_norm": 1.15625,
      "learning_rate": 0.000485710664314899,
      "loss": 5.1505,
      "mean_token_accuracy": 0.19140107333660125,
      "num_tokens": 26967338.0,
      "step": 15545
    },
    {
      "entropy": 5.53289647102356,
      "epoch": 1.2098424431044543,
      "grad_norm": 1.046875,
      "learning_rate": 0.000485700943154485,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20143602043390274,
      "num_tokens": 26976182.0,
      "step": 15550
    },
    {
      "entropy": 5.592840671539307,
      "epoch": 1.210231472476172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004856912187970195,
      "loss": 5.1861,
      "mean_token_accuracy": 0.18550353199243547,
      "num_tokens": 26984275.0,
      "step": 15555
    },
    {
      "entropy": 5.701727437973022,
      "epoch": 1.2106205018478895,
      "grad_norm": 1.125,
      "learning_rate": 0.0004856814912426502,
      "loss": 5.1357,
      "mean_token_accuracy": 0.18943695574998856,
      "num_tokens": 26991891.0,
      "step": 15560
    },
    {
      "entropy": 5.569869089126587,
      "epoch": 1.2110095312196072,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004856717604915245,
      "loss": 5.0576,
      "mean_token_accuracy": 0.1926969826221466,
      "num_tokens": 27000390.0,
      "step": 15565
    },
    {
      "entropy": 5.40335750579834,
      "epoch": 1.2113985605913247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004856620265437902,
      "loss": 5.0511,
      "mean_token_accuracy": 0.20205354392528535,
      "num_tokens": 27009348.0,
      "step": 15570
    },
    {
      "entropy": 5.551998138427734,
      "epoch": 1.2117875899630421,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000485652289399595,
      "loss": 5.1237,
      "mean_token_accuracy": 0.19174213260412215,
      "num_tokens": 27017529.0,
      "step": 15575
    },
    {
      "entropy": 5.628260707855224,
      "epoch": 1.2121766193347598,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048564254905908655,
      "loss": 5.1197,
      "mean_token_accuracy": 0.19197541624307632,
      "num_tokens": 27026238.0,
      "step": 15580
    },
    {
      "entropy": 5.561192798614502,
      "epoch": 1.2125656487064773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048563280552241266,
      "loss": 5.0341,
      "mean_token_accuracy": 0.19890040159225464,
      "num_tokens": 27034995.0,
      "step": 15585
    },
    {
      "entropy": 5.692285966873169,
      "epoch": 1.2129546780781948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004856230587897212,
      "loss": 5.2225,
      "mean_token_accuracy": 0.18564899861812592,
      "num_tokens": 27043631.0,
      "step": 15590
    },
    {
      "entropy": 5.563717699050903,
      "epoch": 1.2133437074499125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004856133088611602,
      "loss": 5.1024,
      "mean_token_accuracy": 0.19418361335992812,
      "num_tokens": 27052491.0,
      "step": 15595
    },
    {
      "entropy": 5.551230430603027,
      "epoch": 1.21373273682163,
      "grad_norm": 1.0,
      "learning_rate": 0.0004856035557368773,
      "loss": 5.0947,
      "mean_token_accuracy": 0.19574228972196578,
      "num_tokens": 27061056.0,
      "step": 15600
    },
    {
      "entropy": 5.503968048095703,
      "epoch": 1.2141217661933477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048559379941702074,
      "loss": 5.0771,
      "mean_token_accuracy": 0.19652830213308334,
      "num_tokens": 27069562.0,
      "step": 15605
    },
    {
      "entropy": 5.597941064834595,
      "epoch": 1.2145107955650651,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048558403990173844,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19686997681856155,
      "num_tokens": 27077814.0,
      "step": 15610
    },
    {
      "entropy": 5.54822564125061,
      "epoch": 1.2148998249367828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048557427719117855,
      "loss": 4.9886,
      "mean_token_accuracy": 0.19760056734085082,
      "num_tokens": 27086515.0,
      "step": 15615
    },
    {
      "entropy": 5.573387813568115,
      "epoch": 1.2152888543085003,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004855645112854891,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19586052745580673,
      "num_tokens": 27095759.0,
      "step": 15620
    },
    {
      "entropy": 5.5309364795684814,
      "epoch": 1.2156778836802178,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004855547421848184,
      "loss": 5.1355,
      "mean_token_accuracy": 0.18665731847286224,
      "num_tokens": 27104921.0,
      "step": 15625
    },
    {
      "entropy": 5.599719858169555,
      "epoch": 1.2160669130519355,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048554496988931456,
      "loss": 5.1082,
      "mean_token_accuracy": 0.19551418721675873,
      "num_tokens": 27113876.0,
      "step": 15630
    },
    {
      "entropy": 5.591343641281128,
      "epoch": 1.216455942423653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048553519439912597,
      "loss": 5.0028,
      "mean_token_accuracy": 0.19517452120780945,
      "num_tokens": 27121654.0,
      "step": 15635
    },
    {
      "entropy": 5.473054313659668,
      "epoch": 1.2168449717953704,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004855254157144009,
      "loss": 5.045,
      "mean_token_accuracy": 0.2030401721596718,
      "num_tokens": 27130156.0,
      "step": 15640
    },
    {
      "entropy": 5.576415586471557,
      "epoch": 1.2172340011670881,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004855156338352877,
      "loss": 5.1462,
      "mean_token_accuracy": 0.18560109436511993,
      "num_tokens": 27138646.0,
      "step": 15645
    },
    {
      "entropy": 5.559660148620606,
      "epoch": 1.2176230305388056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048550584876193493,
      "loss": 5.1131,
      "mean_token_accuracy": 0.194693885743618,
      "num_tokens": 27147282.0,
      "step": 15650
    },
    {
      "entropy": 5.531558609008789,
      "epoch": 1.2180120599105233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048549606049449085,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1818372756242752,
      "num_tokens": 27156238.0,
      "step": 15655
    },
    {
      "entropy": 5.580892515182495,
      "epoch": 1.2184010892822408,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004854862690331041,
      "loss": 5.1317,
      "mean_token_accuracy": 0.19014522135257722,
      "num_tokens": 27164936.0,
      "step": 15660
    },
    {
      "entropy": 5.578609132766724,
      "epoch": 1.2187901186539585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048547647437792327,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1960985079407692,
      "num_tokens": 27173362.0,
      "step": 15665
    },
    {
      "entropy": 5.541981172561646,
      "epoch": 1.219179148025676,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000485466676529097,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1988900139927864,
      "num_tokens": 27181611.0,
      "step": 15670
    },
    {
      "entropy": 5.621472692489624,
      "epoch": 1.2195681773973934,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048545687548677373,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19253577739000322,
      "num_tokens": 27189593.0,
      "step": 15675
    },
    {
      "entropy": 5.572204732894898,
      "epoch": 1.2199572067691111,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004854470712511025,
      "loss": 5.011,
      "mean_token_accuracy": 0.19712709933519362,
      "num_tokens": 27198006.0,
      "step": 15680
    },
    {
      "entropy": 5.594330787658691,
      "epoch": 1.2203462361408286,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048543726382223193,
      "loss": 5.1409,
      "mean_token_accuracy": 0.19214984327554702,
      "num_tokens": 27206846.0,
      "step": 15685
    },
    {
      "entropy": 5.538330221176148,
      "epoch": 1.220735265512546,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048542745320031075,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19127833098173141,
      "num_tokens": 27214838.0,
      "step": 15690
    },
    {
      "entropy": 5.576240921020508,
      "epoch": 1.2211242948842638,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048541763938548795,
      "loss": 5.1629,
      "mean_token_accuracy": 0.18567192703485488,
      "num_tokens": 27224161.0,
      "step": 15695
    },
    {
      "entropy": 5.534440994262695,
      "epoch": 1.2215133242559812,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048540782237791244,
      "loss": 5.0239,
      "mean_token_accuracy": 0.18766839057207108,
      "num_tokens": 27233084.0,
      "step": 15700
    },
    {
      "entropy": 5.6995758533477785,
      "epoch": 1.221902353627699,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048539800217773293,
      "loss": 5.2129,
      "mean_token_accuracy": 0.181914059817791,
      "num_tokens": 27242219.0,
      "step": 15705
    },
    {
      "entropy": 5.63449912071228,
      "epoch": 1.2222913829994164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004853881787850988,
      "loss": 5.1671,
      "mean_token_accuracy": 0.1890419602394104,
      "num_tokens": 27250411.0,
      "step": 15710
    },
    {
      "entropy": 5.648552274703979,
      "epoch": 1.2226804123711341,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048537835220015886,
      "loss": 5.173,
      "mean_token_accuracy": 0.184712253510952,
      "num_tokens": 27259472.0,
      "step": 15715
    },
    {
      "entropy": 5.587137842178345,
      "epoch": 1.2230694417428516,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004853685224230623,
      "loss": 5.0634,
      "mean_token_accuracy": 0.19516150802373886,
      "num_tokens": 27267922.0,
      "step": 15720
    },
    {
      "entropy": 5.56849856376648,
      "epoch": 1.223458471114569,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048535868945395825,
      "loss": 5.1578,
      "mean_token_accuracy": 0.19511961042881013,
      "num_tokens": 27277129.0,
      "step": 15725
    },
    {
      "entropy": 5.638340091705322,
      "epoch": 1.2238475004862868,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048534885329299586,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19260670989751816,
      "num_tokens": 27285583.0,
      "step": 15730
    },
    {
      "entropy": 5.566179418563843,
      "epoch": 1.2242365298580042,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853390139403245,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1962965965270996,
      "num_tokens": 27294292.0,
      "step": 15735
    },
    {
      "entropy": 5.608720350265503,
      "epoch": 1.224625559229722,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048532917139609334,
      "loss": 5.2304,
      "mean_token_accuracy": 0.1838957503437996,
      "num_tokens": 27302606.0,
      "step": 15740
    },
    {
      "entropy": 5.645279169082642,
      "epoch": 1.2250145886014394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004853193256604518,
      "loss": 5.0819,
      "mean_token_accuracy": 0.1970806285738945,
      "num_tokens": 27311103.0,
      "step": 15745
    },
    {
      "entropy": 5.658607625961304,
      "epoch": 1.2254036179731569,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048530947673354924,
      "loss": 5.1,
      "mean_token_accuracy": 0.19465262293815613,
      "num_tokens": 27319497.0,
      "step": 15750
    },
    {
      "entropy": 5.59012131690979,
      "epoch": 1.2257926473448746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004852996246155351,
      "loss": 5.1536,
      "mean_token_accuracy": 0.18796009868383406,
      "num_tokens": 27328367.0,
      "step": 15755
    },
    {
      "entropy": 5.559670305252075,
      "epoch": 1.226181676716592,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048528976930655896,
      "loss": 5.0405,
      "mean_token_accuracy": 0.19445709139108658,
      "num_tokens": 27336797.0,
      "step": 15760
    },
    {
      "entropy": 5.519541835784912,
      "epoch": 1.2265707060883098,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048527991080677015,
      "loss": 5.0104,
      "mean_token_accuracy": 0.1938612252473831,
      "num_tokens": 27345052.0,
      "step": 15765
    },
    {
      "entropy": 5.577302026748657,
      "epoch": 1.2269597354600272,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048527004911631843,
      "loss": 5.0523,
      "mean_token_accuracy": 0.1978906974196434,
      "num_tokens": 27353062.0,
      "step": 15770
    },
    {
      "entropy": 5.582893896102905,
      "epoch": 1.2273487648317447,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004852601842353534,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19258177131414414,
      "num_tokens": 27361471.0,
      "step": 15775
    },
    {
      "entropy": 5.658461713790894,
      "epoch": 1.2277377942034624,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048525031616402476,
      "loss": 5.1533,
      "mean_token_accuracy": 0.18988002836704254,
      "num_tokens": 27370655.0,
      "step": 15780
    },
    {
      "entropy": 5.5359241485595705,
      "epoch": 1.2281268235751799,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048524044490248227,
      "loss": 4.9843,
      "mean_token_accuracy": 0.20162599682807922,
      "num_tokens": 27379329.0,
      "step": 15785
    },
    {
      "entropy": 5.614940595626831,
      "epoch": 1.2285158529468976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048523057045087557,
      "loss": 5.12,
      "mean_token_accuracy": 0.19115816205739974,
      "num_tokens": 27388864.0,
      "step": 15790
    },
    {
      "entropy": 5.646668910980225,
      "epoch": 1.228904882318615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048522069280935466,
      "loss": 5.232,
      "mean_token_accuracy": 0.173755244910717,
      "num_tokens": 27397804.0,
      "step": 15795
    },
    {
      "entropy": 5.575143194198608,
      "epoch": 1.2292939116903325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004852108119780693,
      "loss": 5.0846,
      "mean_token_accuracy": 0.1886795222759247,
      "num_tokens": 27406578.0,
      "step": 15800
    },
    {
      "entropy": 5.623072910308838,
      "epoch": 1.2296829410620502,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004852009279571694,
      "loss": 5.1818,
      "mean_token_accuracy": 0.18230880796909332,
      "num_tokens": 27415207.0,
      "step": 15805
    },
    {
      "entropy": 5.594999885559082,
      "epoch": 1.2300719704337677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000485191040746805,
      "loss": 5.1059,
      "mean_token_accuracy": 0.1887709379196167,
      "num_tokens": 27424201.0,
      "step": 15810
    },
    {
      "entropy": 5.569968128204346,
      "epoch": 1.2304609998054854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004851811503471262,
      "loss": 5.0891,
      "mean_token_accuracy": 0.19137098640203476,
      "num_tokens": 27433187.0,
      "step": 15815
    },
    {
      "entropy": 5.554637622833252,
      "epoch": 1.2308500291772029,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048517125675828294,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19239335656166076,
      "num_tokens": 27441976.0,
      "step": 15820
    },
    {
      "entropy": 5.549738645553589,
      "epoch": 1.2312390585489204,
      "grad_norm": 1.125,
      "learning_rate": 0.00048516135998042536,
      "loss": 5.0301,
      "mean_token_accuracy": 0.19385307729244233,
      "num_tokens": 27450235.0,
      "step": 15825
    },
    {
      "entropy": 5.6378621578216555,
      "epoch": 1.231628087920638,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004851514600137037,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18178775459527968,
      "num_tokens": 27458842.0,
      "step": 15830
    },
    {
      "entropy": 5.654992294311524,
      "epoch": 1.2320171172923555,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048514155685826807,
      "loss": 5.169,
      "mean_token_accuracy": 0.19159705340862274,
      "num_tokens": 27467383.0,
      "step": 15835
    },
    {
      "entropy": 5.566771936416626,
      "epoch": 1.2324061466640732,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004851316505142688,
      "loss": 5.1354,
      "mean_token_accuracy": 0.191806398332119,
      "num_tokens": 27475811.0,
      "step": 15840
    },
    {
      "entropy": 5.663056421279907,
      "epoch": 1.2327951760357907,
      "grad_norm": 1.125,
      "learning_rate": 0.00048512174098185615,
      "loss": 5.2413,
      "mean_token_accuracy": 0.17960155308246611,
      "num_tokens": 27484032.0,
      "step": 15845
    },
    {
      "entropy": 5.560014295578003,
      "epoch": 1.2331842054075082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048511182826118055,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2033701166510582,
      "num_tokens": 27492330.0,
      "step": 15850
    },
    {
      "entropy": 5.487945032119751,
      "epoch": 1.2335732347792259,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048510191235239246,
      "loss": 5.0786,
      "mean_token_accuracy": 0.19590992629528045,
      "num_tokens": 27501042.0,
      "step": 15855
    },
    {
      "entropy": 5.552310991287231,
      "epoch": 1.2339622641509433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048509199325564217,
      "loss": 5.0219,
      "mean_token_accuracy": 0.19686900526285173,
      "num_tokens": 27509939.0,
      "step": 15860
    },
    {
      "entropy": 5.627854299545288,
      "epoch": 1.234351293522661,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004850820709710803,
      "loss": 5.2237,
      "mean_token_accuracy": 0.17911992967128754,
      "num_tokens": 27518653.0,
      "step": 15865
    },
    {
      "entropy": 5.600654792785645,
      "epoch": 1.2347403228943785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004850721454988574,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19784297794103622,
      "num_tokens": 27526856.0,
      "step": 15870
    },
    {
      "entropy": 5.53119421005249,
      "epoch": 1.235129352266096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048506221683912405,
      "loss": 5.0245,
      "mean_token_accuracy": 0.19851939678192138,
      "num_tokens": 27535205.0,
      "step": 15875
    },
    {
      "entropy": 5.611048316955566,
      "epoch": 1.2355183816378137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004850522849920309,
      "loss": 5.118,
      "mean_token_accuracy": 0.19082388728857042,
      "num_tokens": 27543516.0,
      "step": 15880
    },
    {
      "entropy": 5.626580905914307,
      "epoch": 1.2359074110095312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048504234995772863,
      "loss": 5.1564,
      "mean_token_accuracy": 0.18912261575460435,
      "num_tokens": 27552136.0,
      "step": 15885
    },
    {
      "entropy": 5.550416564941406,
      "epoch": 1.2362964403812489,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004850324117363681,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20084468573331832,
      "num_tokens": 27560805.0,
      "step": 15890
    },
    {
      "entropy": 5.531932401657104,
      "epoch": 1.2366854697529663,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048502247032809997,
      "loss": 5.0654,
      "mean_token_accuracy": 0.19469710439443588,
      "num_tokens": 27569643.0,
      "step": 15895
    },
    {
      "entropy": 5.569939517974854,
      "epoch": 1.237074499124684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004850125257330751,
      "loss": 5.014,
      "mean_token_accuracy": 0.19518553763628005,
      "num_tokens": 27578506.0,
      "step": 15900
    },
    {
      "entropy": 5.53627028465271,
      "epoch": 1.2374635284964015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048500257795144446,
      "loss": 5.1064,
      "mean_token_accuracy": 0.1886509507894516,
      "num_tokens": 27586911.0,
      "step": 15905
    },
    {
      "entropy": 5.581449699401856,
      "epoch": 1.237852557868119,
      "grad_norm": 1.0,
      "learning_rate": 0.000484992626983359,
      "loss": 5.1529,
      "mean_token_accuracy": 0.18771650940179824,
      "num_tokens": 27595112.0,
      "step": 15910
    },
    {
      "entropy": 5.622104263305664,
      "epoch": 1.2382415872398367,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004849826728289696,
      "loss": 5.1515,
      "mean_token_accuracy": 0.18373007327318192,
      "num_tokens": 27603923.0,
      "step": 15915
    },
    {
      "entropy": 5.5163239479064945,
      "epoch": 1.2386306166115542,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048497271548842737,
      "loss": 4.9777,
      "mean_token_accuracy": 0.20386564880609512,
      "num_tokens": 27612705.0,
      "step": 15920
    },
    {
      "entropy": 5.548664712905884,
      "epoch": 1.2390196459832716,
      "grad_norm": 1.0,
      "learning_rate": 0.0004849627549618834,
      "loss": 5.0832,
      "mean_token_accuracy": 0.19662585854530334,
      "num_tokens": 27621858.0,
      "step": 15925
    },
    {
      "entropy": 5.630874967575073,
      "epoch": 1.2394086753549893,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048495279124948886,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1876761272549629,
      "num_tokens": 27629977.0,
      "step": 15930
    },
    {
      "entropy": 5.640757465362549,
      "epoch": 1.2397977047267068,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004849428243513948,
      "loss": 5.127,
      "mean_token_accuracy": 0.19379250407218934,
      "num_tokens": 27638433.0,
      "step": 15935
    },
    {
      "entropy": 5.541794204711914,
      "epoch": 1.2401867340984245,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048493285426775255,
      "loss": 5.0331,
      "mean_token_accuracy": 0.19448391497135162,
      "num_tokens": 27647116.0,
      "step": 15940
    },
    {
      "entropy": 5.535487937927246,
      "epoch": 1.240575763470142,
      "grad_norm": 1.125,
      "learning_rate": 0.0004849228809987135,
      "loss": 5.115,
      "mean_token_accuracy": 0.18874460756778716,
      "num_tokens": 27655598.0,
      "step": 15945
    },
    {
      "entropy": 5.652612733840942,
      "epoch": 1.2409647928418597,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004849129045444288,
      "loss": 5.2059,
      "mean_token_accuracy": 0.1868632107973099,
      "num_tokens": 27664308.0,
      "step": 15950
    },
    {
      "entropy": 5.528329277038575,
      "epoch": 1.2413538222135772,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004849029249050498,
      "loss": 4.9597,
      "mean_token_accuracy": 0.20222141593694687,
      "num_tokens": 27672808.0,
      "step": 15955
    },
    {
      "entropy": 5.604690790176392,
      "epoch": 1.2417428515852946,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048489294208072805,
      "loss": 5.1912,
      "mean_token_accuracy": 0.18527125865221022,
      "num_tokens": 27681091.0,
      "step": 15960
    },
    {
      "entropy": 5.568545341491699,
      "epoch": 1.2421318809570123,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048488295607161495,
      "loss": 5.1132,
      "mean_token_accuracy": 0.19552708864212037,
      "num_tokens": 27689582.0,
      "step": 15965
    },
    {
      "entropy": 5.560872316360474,
      "epoch": 1.2425209103287298,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004848729668778621,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19819256365299226,
      "num_tokens": 27698509.0,
      "step": 15970
    },
    {
      "entropy": 5.606788873672485,
      "epoch": 1.2429099397004473,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004848629744996211,
      "loss": 5.1492,
      "mean_token_accuracy": 0.18455335795879363,
      "num_tokens": 27708216.0,
      "step": 15975
    },
    {
      "entropy": 5.6683145523071286,
      "epoch": 1.243298969072165,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004848529789370434,
      "loss": 5.1589,
      "mean_token_accuracy": 0.18668816536664962,
      "num_tokens": 27716801.0,
      "step": 15980
    },
    {
      "entropy": 5.591026401519775,
      "epoch": 1.2436879984438824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004848429801902808,
      "loss": 5.1184,
      "mean_token_accuracy": 0.19335718154907228,
      "num_tokens": 27725186.0,
      "step": 15985
    },
    {
      "entropy": 5.65595440864563,
      "epoch": 1.2440770278156001,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048483297825948485,
      "loss": 5.079,
      "mean_token_accuracy": 0.19125808626413346,
      "num_tokens": 27733386.0,
      "step": 15990
    },
    {
      "entropy": 5.577162456512451,
      "epoch": 1.2444660571873176,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048482297314480756,
      "loss": 5.134,
      "mean_token_accuracy": 0.18923188000917435,
      "num_tokens": 27742253.0,
      "step": 15995
    },
    {
      "entropy": 5.543339109420776,
      "epoch": 1.2448550865590353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004848129648464006,
      "loss": 5.0574,
      "mean_token_accuracy": 0.19739408046007156,
      "num_tokens": 27751555.0,
      "step": 16000
    },
    {
      "entropy": 5.625134134292603,
      "epoch": 1.2452441159307528,
      "grad_norm": 1.125,
      "learning_rate": 0.0004848029533644159,
      "loss": 5.2063,
      "mean_token_accuracy": 0.18494432121515275,
      "num_tokens": 27760151.0,
      "step": 16005
    },
    {
      "entropy": 5.6406434059143065,
      "epoch": 1.2456331453024703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004847929386990052,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19732231199741362,
      "num_tokens": 27770017.0,
      "step": 16010
    },
    {
      "entropy": 5.569770812988281,
      "epoch": 1.246022174674188,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048478292085032065,
      "loss": 5.0098,
      "mean_token_accuracy": 0.19767328649759291,
      "num_tokens": 27777796.0,
      "step": 16015
    },
    {
      "entropy": 5.529715204238892,
      "epoch": 1.2464112040459054,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004847728998185142,
      "loss": 5.1267,
      "mean_token_accuracy": 0.19125607162714003,
      "num_tokens": 27785784.0,
      "step": 16020
    },
    {
      "entropy": 5.637019205093384,
      "epoch": 1.246800233417623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048476287560373786,
      "loss": 5.0904,
      "mean_token_accuracy": 0.18956329971551894,
      "num_tokens": 27793911.0,
      "step": 16025
    },
    {
      "entropy": 5.553223896026611,
      "epoch": 1.2471892627893406,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048475284820614375,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19351170659065248,
      "num_tokens": 27803125.0,
      "step": 16030
    },
    {
      "entropy": 5.530981492996216,
      "epoch": 1.247578292161058,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048474281762588407,
      "loss": 5.0695,
      "mean_token_accuracy": 0.196527361869812,
      "num_tokens": 27812008.0,
      "step": 16035
    },
    {
      "entropy": 5.60683274269104,
      "epoch": 1.2479673215327758,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004847327838631109,
      "loss": 5.1915,
      "mean_token_accuracy": 0.188255412876606,
      "num_tokens": 27820490.0,
      "step": 16040
    },
    {
      "entropy": 5.684503126144409,
      "epoch": 1.2483563509044933,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004847227469179766,
      "loss": 5.1391,
      "mean_token_accuracy": 0.19031050205230712,
      "num_tokens": 27829513.0,
      "step": 16045
    },
    {
      "entropy": 5.595043802261353,
      "epoch": 1.248745380276211,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004847127067906334,
      "loss": 5.0602,
      "mean_token_accuracy": 0.19469755738973618,
      "num_tokens": 27837811.0,
      "step": 16050
    },
    {
      "entropy": 5.628044080734253,
      "epoch": 1.2491344096479284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048470266348123365,
      "loss": 5.2176,
      "mean_token_accuracy": 0.18309867084026338,
      "num_tokens": 27846449.0,
      "step": 16055
    },
    {
      "entropy": 5.558897066116333,
      "epoch": 1.249523439019646,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004846926169899298,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2016116514801979,
      "num_tokens": 27854834.0,
      "step": 16060
    },
    {
      "entropy": 5.634098100662231,
      "epoch": 1.2499124683913636,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048468256731687426,
      "loss": 5.1498,
      "mean_token_accuracy": 0.1885363668203354,
      "num_tokens": 27862984.0,
      "step": 16065
    },
    {
      "entropy": 5.662493371963501,
      "epoch": 1.250301497763081,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004846725144622194,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1834780380129814,
      "num_tokens": 27872198.0,
      "step": 16070
    },
    {
      "entropy": 5.666763591766357,
      "epoch": 1.2506905271347986,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000484662458426118,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1958122655749321,
      "num_tokens": 27880856.0,
      "step": 16075
    },
    {
      "entropy": 5.534097480773926,
      "epoch": 1.2510795565065163,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048465239920872247,
      "loss": 5.02,
      "mean_token_accuracy": 0.20026249885559083,
      "num_tokens": 27888677.0,
      "step": 16080
    },
    {
      "entropy": 5.531265020370483,
      "epoch": 1.2514685858782337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048464233681018545,
      "loss": 5.0854,
      "mean_token_accuracy": 0.19213679283857346,
      "num_tokens": 27896940.0,
      "step": 16085
    },
    {
      "entropy": 5.618250942230224,
      "epoch": 1.2518576152499514,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004846322712306596,
      "loss": 5.1601,
      "mean_token_accuracy": 0.19110050648450852,
      "num_tokens": 27905739.0,
      "step": 16090
    },
    {
      "entropy": 5.64260458946228,
      "epoch": 1.252246644621669,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048462220247029783,
      "loss": 5.1209,
      "mean_token_accuracy": 0.19774194806814194,
      "num_tokens": 27914029.0,
      "step": 16095
    },
    {
      "entropy": 5.624641275405883,
      "epoch": 1.2526356739933866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048461213052925265,
      "loss": 5.2257,
      "mean_token_accuracy": 0.18462551832199098,
      "num_tokens": 27923287.0,
      "step": 16100
    },
    {
      "entropy": 5.614475059509277,
      "epoch": 1.253024703365104,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004846020554076771,
      "loss": 5.0866,
      "mean_token_accuracy": 0.190290929377079,
      "num_tokens": 27931727.0,
      "step": 16105
    },
    {
      "entropy": 5.572029304504395,
      "epoch": 1.2534137327368216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004845919771057239,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2002420797944069,
      "num_tokens": 27939524.0,
      "step": 16110
    },
    {
      "entropy": 5.589701175689697,
      "epoch": 1.2538027621085392,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845818956235462,
      "loss": 5.0866,
      "mean_token_accuracy": 0.19052526205778123,
      "num_tokens": 27947827.0,
      "step": 16115
    },
    {
      "entropy": 5.510835742950439,
      "epoch": 1.2541917914802567,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048457181096129664,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19661034196615218,
      "num_tokens": 27956447.0,
      "step": 16120
    },
    {
      "entropy": 5.619442892074585,
      "epoch": 1.2545808208519742,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004845617231191285,
      "loss": 5.1416,
      "mean_token_accuracy": 0.18829835951328278,
      "num_tokens": 27965570.0,
      "step": 16125
    },
    {
      "entropy": 5.570482873916626,
      "epoch": 1.254969850223692,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004845516320971948,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1849375456571579,
      "num_tokens": 27974527.0,
      "step": 16130
    },
    {
      "entropy": 5.528414344787597,
      "epoch": 1.2553588795954094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004845415378956486,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19365038722753525,
      "num_tokens": 27983190.0,
      "step": 16135
    },
    {
      "entropy": 5.5938920974731445,
      "epoch": 1.255747908967127,
      "grad_norm": 1.140625,
      "learning_rate": 0.000484531440514643,
      "loss": 5.1485,
      "mean_token_accuracy": 0.18940896093845366,
      "num_tokens": 27992203.0,
      "step": 16140
    },
    {
      "entropy": 5.615431308746338,
      "epoch": 1.2561369383388445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048452133995433136,
      "loss": 5.0978,
      "mean_token_accuracy": 0.19469057619571686,
      "num_tokens": 28001154.0,
      "step": 16145
    },
    {
      "entropy": 5.5246817111969,
      "epoch": 1.2565259677105622,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004845112362148668,
      "loss": 4.9991,
      "mean_token_accuracy": 0.1956571266055107,
      "num_tokens": 28009841.0,
      "step": 16150
    },
    {
      "entropy": 5.536205148696899,
      "epoch": 1.2569149970822797,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004845011292964028,
      "loss": 5.1048,
      "mean_token_accuracy": 0.19138651043176652,
      "num_tokens": 28019245.0,
      "step": 16155
    },
    {
      "entropy": 5.698300552368164,
      "epoch": 1.2573040264539972,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048449101919909265,
      "loss": 5.2214,
      "mean_token_accuracy": 0.18296273946762084,
      "num_tokens": 28028342.0,
      "step": 16160
    },
    {
      "entropy": 5.672586679458618,
      "epoch": 1.257693055825715,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048448090592308955,
      "loss": 5.1701,
      "mean_token_accuracy": 0.18794131726026536,
      "num_tokens": 28037308.0,
      "step": 16165
    },
    {
      "entropy": 5.623469066619873,
      "epoch": 1.2580820851974324,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004844707894685473,
      "loss": 5.1467,
      "mean_token_accuracy": 0.19212767779827117,
      "num_tokens": 28046214.0,
      "step": 16170
    },
    {
      "entropy": 5.536723279953003,
      "epoch": 1.2584711145691498,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004844606698356192,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2015048608183861,
      "num_tokens": 28054722.0,
      "step": 16175
    },
    {
      "entropy": 5.537246990203857,
      "epoch": 1.2588601439408675,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004844505470244588,
      "loss": 4.9933,
      "mean_token_accuracy": 0.19242886304855347,
      "num_tokens": 28062596.0,
      "step": 16180
    },
    {
      "entropy": 5.503886699676514,
      "epoch": 1.2592491733125852,
      "grad_norm": 1.125,
      "learning_rate": 0.0004844404210352197,
      "loss": 5.0358,
      "mean_token_accuracy": 0.19412676393985748,
      "num_tokens": 28071183.0,
      "step": 16185
    },
    {
      "entropy": 5.502623748779297,
      "epoch": 1.2596382026843027,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048443029186805554,
      "loss": 5.0369,
      "mean_token_accuracy": 0.19789378345012665,
      "num_tokens": 28079194.0,
      "step": 16190
    },
    {
      "entropy": 5.50315146446228,
      "epoch": 1.2600272320560202,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004844201595231201,
      "loss": 5.0398,
      "mean_token_accuracy": 0.19202157557010652,
      "num_tokens": 28087874.0,
      "step": 16195
    },
    {
      "entropy": 5.661500024795532,
      "epoch": 1.2604162614277379,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048441002400056706,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1928700953722,
      "num_tokens": 28097390.0,
      "step": 16200
    },
    {
      "entropy": 5.547211790084839,
      "epoch": 1.2608052907994554,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004843998853005502,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19320148080587388,
      "num_tokens": 28105445.0,
      "step": 16205
    },
    {
      "entropy": 5.594481801986694,
      "epoch": 1.2611943201711728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004843897434232233,
      "loss": 5.0654,
      "mean_token_accuracy": 0.1963140457868576,
      "num_tokens": 28113855.0,
      "step": 16210
    },
    {
      "entropy": 5.564957094192505,
      "epoch": 1.2615833495428905,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048437959836874047,
      "loss": 5.0327,
      "mean_token_accuracy": 0.19826897382736205,
      "num_tokens": 28123300.0,
      "step": 16215
    },
    {
      "entropy": 5.546922397613526,
      "epoch": 1.261972378914608,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048436945013725544,
      "loss": 5.0228,
      "mean_token_accuracy": 0.19342950731515884,
      "num_tokens": 28131829.0,
      "step": 16220
    },
    {
      "entropy": 5.593236970901489,
      "epoch": 1.2623614082863255,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048435929872892226,
      "loss": 5.1522,
      "mean_token_accuracy": 0.18728365004062653,
      "num_tokens": 28141060.0,
      "step": 16225
    },
    {
      "entropy": 5.6218953132629395,
      "epoch": 1.2627504376580432,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004843491441438949,
      "loss": 5.1717,
      "mean_token_accuracy": 0.1858040750026703,
      "num_tokens": 28149866.0,
      "step": 16230
    },
    {
      "entropy": 5.578256893157959,
      "epoch": 1.2631394670297609,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048433898638232755,
      "loss": 5.1232,
      "mean_token_accuracy": 0.18676017969846725,
      "num_tokens": 28158593.0,
      "step": 16235
    },
    {
      "entropy": 5.574200201034546,
      "epoch": 1.2635284964014784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004843288254443742,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1999283567070961,
      "num_tokens": 28167301.0,
      "step": 16240
    },
    {
      "entropy": 5.571263599395752,
      "epoch": 1.2639175257731958,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004843186613301892,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19715780317783355,
      "num_tokens": 28176092.0,
      "step": 16245
    },
    {
      "entropy": 5.66141676902771,
      "epoch": 1.2643065551449135,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004843084940399266,
      "loss": 5.175,
      "mean_token_accuracy": 0.19590607434511184,
      "num_tokens": 28184570.0,
      "step": 16250
    },
    {
      "entropy": 5.584057903289795,
      "epoch": 1.264695584516631,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004842983235737408,
      "loss": 5.0925,
      "mean_token_accuracy": 0.1999397173523903,
      "num_tokens": 28192874.0,
      "step": 16255
    },
    {
      "entropy": 5.605925559997559,
      "epoch": 1.2650846138883485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004842881499317861,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19828723669052123,
      "num_tokens": 28201546.0,
      "step": 16260
    },
    {
      "entropy": 5.58402099609375,
      "epoch": 1.2654736432600662,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004842779731142168,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1916880339384079,
      "num_tokens": 28210061.0,
      "step": 16265
    },
    {
      "entropy": 5.540750551223755,
      "epoch": 1.2658626726317836,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048426779312118735,
      "loss": 5.0158,
      "mean_token_accuracy": 0.19447221606969833,
      "num_tokens": 28219334.0,
      "step": 16270
    },
    {
      "entropy": 5.663887596130371,
      "epoch": 1.2662517020035013,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004842576099528522,
      "loss": 5.1396,
      "mean_token_accuracy": 0.18783161491155625,
      "num_tokens": 28227963.0,
      "step": 16275
    },
    {
      "entropy": 5.569371843338013,
      "epoch": 1.2666407313752188,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004842474236093659,
      "loss": 5.1069,
      "mean_token_accuracy": 0.190003801882267,
      "num_tokens": 28236648.0,
      "step": 16280
    },
    {
      "entropy": 5.564645385742187,
      "epoch": 1.2670297607469365,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048423723409088306,
      "loss": 5.0965,
      "mean_token_accuracy": 0.19149112105369567,
      "num_tokens": 28245221.0,
      "step": 16285
    },
    {
      "entropy": 5.554838371276856,
      "epoch": 1.267418790118654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004842270413975582,
      "loss": 5.1035,
      "mean_token_accuracy": 0.18969713747501374,
      "num_tokens": 28254481.0,
      "step": 16290
    },
    {
      "entropy": 5.623555421829224,
      "epoch": 1.2678078194903715,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000484216845529546,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19377092123031617,
      "num_tokens": 28263844.0,
      "step": 16295
    },
    {
      "entropy": 5.562657451629638,
      "epoch": 1.2681968488620892,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048420664648700113,
      "loss": 5.0471,
      "mean_token_accuracy": 0.18926697522401809,
      "num_tokens": 28272500.0,
      "step": 16300
    },
    {
      "entropy": 5.629384756088257,
      "epoch": 1.2685858782338066,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004841964442700784,
      "loss": 5.1101,
      "mean_token_accuracy": 0.18592654317617416,
      "num_tokens": 28281553.0,
      "step": 16305
    },
    {
      "entropy": 5.6003162384033205,
      "epoch": 1.2689749076055241,
      "grad_norm": 1.125,
      "learning_rate": 0.00048418623887893264,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19503989517688752,
      "num_tokens": 28290078.0,
      "step": 16310
    },
    {
      "entropy": 5.524891757965088,
      "epoch": 1.2693639369772418,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048417603031371867,
      "loss": 4.9348,
      "mean_token_accuracy": 0.19826858043670653,
      "num_tokens": 28298879.0,
      "step": 16315
    },
    {
      "entropy": 5.525608539581299,
      "epoch": 1.2697529663489593,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004841658185745913,
      "loss": 5.0539,
      "mean_token_accuracy": 0.1975033164024353,
      "num_tokens": 28307188.0,
      "step": 16320
    },
    {
      "entropy": 5.520714330673218,
      "epoch": 1.270141995720677,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048415560366170564,
      "loss": 5.0136,
      "mean_token_accuracy": 0.19913490265607833,
      "num_tokens": 28316012.0,
      "step": 16325
    },
    {
      "entropy": 5.605494785308838,
      "epoch": 1.2705310250923945,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004841453855752165,
      "loss": 5.1726,
      "mean_token_accuracy": 0.1877582550048828,
      "num_tokens": 28325750.0,
      "step": 16330
    },
    {
      "entropy": 5.549591016769409,
      "epoch": 1.2709200544641122,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004841351643152791,
      "loss": 5.1444,
      "mean_token_accuracy": 0.18847979009151458,
      "num_tokens": 28335217.0,
      "step": 16335
    },
    {
      "entropy": 5.610859537124634,
      "epoch": 1.2713090838358296,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004841249398820485,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19685233384370804,
      "num_tokens": 28344123.0,
      "step": 16340
    },
    {
      "entropy": 5.608767795562744,
      "epoch": 1.271698113207547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004841147122756797,
      "loss": 5.0936,
      "mean_token_accuracy": 0.18979732245206832,
      "num_tokens": 28353230.0,
      "step": 16345
    },
    {
      "entropy": 5.588328886032104,
      "epoch": 1.2720871425792648,
      "grad_norm": 1.109375,
      "learning_rate": 0.000484104481496328,
      "loss": 5.1269,
      "mean_token_accuracy": 0.1860384836792946,
      "num_tokens": 28361636.0,
      "step": 16350
    },
    {
      "entropy": 5.561887216567993,
      "epoch": 1.2724761719509823,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004840942475441486,
      "loss": 5.069,
      "mean_token_accuracy": 0.18840545415878296,
      "num_tokens": 28370066.0,
      "step": 16355
    },
    {
      "entropy": 5.571313810348511,
      "epoch": 1.2728652013226998,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004840840104192969,
      "loss": 5.1496,
      "mean_token_accuracy": 0.1873210310935974,
      "num_tokens": 28379017.0,
      "step": 16360
    },
    {
      "entropy": 5.6400126934051515,
      "epoch": 1.2732542306944175,
      "grad_norm": 1.03125,
      "learning_rate": 0.000484073770121928,
      "loss": 5.1142,
      "mean_token_accuracy": 0.19125721901655196,
      "num_tokens": 28387939.0,
      "step": 16365
    },
    {
      "entropy": 5.553504276275635,
      "epoch": 1.273643260066135,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004840635266521975,
      "loss": 5.0094,
      "mean_token_accuracy": 0.19649734646081923,
      "num_tokens": 28396341.0,
      "step": 16370
    },
    {
      "entropy": 5.529841041564941,
      "epoch": 1.2740322894378526,
      "grad_norm": 1.125,
      "learning_rate": 0.0004840532800102607,
      "loss": 5.0446,
      "mean_token_accuracy": 0.19250058829784394,
      "num_tokens": 28405128.0,
      "step": 16375
    },
    {
      "entropy": 5.554614925384522,
      "epoch": 1.27442131880957,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048404303019627314,
      "loss": 5.0396,
      "mean_token_accuracy": 0.1956628143787384,
      "num_tokens": 28412841.0,
      "step": 16380
    },
    {
      "entropy": 5.538470888137818,
      "epoch": 1.2748103481812878,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048403277721039036,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20153042823076248,
      "num_tokens": 28421526.0,
      "step": 16385
    },
    {
      "entropy": 5.560793018341064,
      "epoch": 1.2751993775530053,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004840225210527679,
      "loss": 5.1479,
      "mean_token_accuracy": 0.18904553353786469,
      "num_tokens": 28430070.0,
      "step": 16390
    },
    {
      "entropy": 5.50371618270874,
      "epoch": 1.2755884069247228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004840122617235613,
      "loss": 5.0138,
      "mean_token_accuracy": 0.19572129100561142,
      "num_tokens": 28438920.0,
      "step": 16395
    },
    {
      "entropy": 5.589546012878418,
      "epoch": 1.2759774362964404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004840019992229263,
      "loss": 5.0717,
      "mean_token_accuracy": 0.19732384234666825,
      "num_tokens": 28447666.0,
      "step": 16400
    },
    {
      "entropy": 5.585099458694458,
      "epoch": 1.276366465668158,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048399173355101867,
      "loss": 5.18,
      "mean_token_accuracy": 0.18551851361989974,
      "num_tokens": 28457341.0,
      "step": 16405
    },
    {
      "entropy": 5.564286136627198,
      "epoch": 1.2767554950398754,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048398146470799417,
      "loss": 5.0663,
      "mean_token_accuracy": 0.2015332892537117,
      "num_tokens": 28466202.0,
      "step": 16410
    },
    {
      "entropy": 5.6681740283966064,
      "epoch": 1.277144524411593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048397119269400846,
      "loss": 5.1646,
      "mean_token_accuracy": 0.1919760912656784,
      "num_tokens": 28474494.0,
      "step": 16415
    },
    {
      "entropy": 5.609622573852539,
      "epoch": 1.2775335537833106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048396091750921766,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19252357631921768,
      "num_tokens": 28482931.0,
      "step": 16420
    },
    {
      "entropy": 5.561962842941284,
      "epoch": 1.2779225831550283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004839506391537774,
      "loss": 5.081,
      "mean_token_accuracy": 0.1932113617658615,
      "num_tokens": 28491372.0,
      "step": 16425
    },
    {
      "entropy": 5.520247220993042,
      "epoch": 1.2783116125267457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004839403576278438,
      "loss": 4.9953,
      "mean_token_accuracy": 0.19384261071681977,
      "num_tokens": 28500201.0,
      "step": 16430
    },
    {
      "entropy": 5.556700134277344,
      "epoch": 1.2787006418984634,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004839300729315729,
      "loss": 5.0345,
      "mean_token_accuracy": 0.19640935957431793,
      "num_tokens": 28508899.0,
      "step": 16435
    },
    {
      "entropy": 5.547556066513062,
      "epoch": 1.279089671270181,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004839197850651206,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19372555166482924,
      "num_tokens": 28518280.0,
      "step": 16440
    },
    {
      "entropy": 5.619271373748779,
      "epoch": 1.2794787006418984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048390949402864317,
      "loss": 5.1551,
      "mean_token_accuracy": 0.18661375641822814,
      "num_tokens": 28527113.0,
      "step": 16445
    },
    {
      "entropy": 5.537773513793946,
      "epoch": 1.279867730013616,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048389919982229666,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19924020171165466,
      "num_tokens": 28535253.0,
      "step": 16450
    },
    {
      "entropy": 5.517531061172486,
      "epoch": 1.2802567593853336,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048388890244623723,
      "loss": 5.0988,
      "mean_token_accuracy": 0.1910899430513382,
      "num_tokens": 28543475.0,
      "step": 16455
    },
    {
      "entropy": 5.607239055633545,
      "epoch": 1.280645788757051,
      "grad_norm": 1.125,
      "learning_rate": 0.0004838786019006213,
      "loss": 5.1488,
      "mean_token_accuracy": 0.19588906764984132,
      "num_tokens": 28551921.0,
      "step": 16460
    },
    {
      "entropy": 5.595171976089477,
      "epoch": 1.2810348181287687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048386829818560493,
      "loss": 5.1196,
      "mean_token_accuracy": 0.19057063460350038,
      "num_tokens": 28560413.0,
      "step": 16465
    },
    {
      "entropy": 5.6245489597320555,
      "epoch": 1.2814238475004862,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048385799130134464,
      "loss": 5.1528,
      "mean_token_accuracy": 0.19450900852680206,
      "num_tokens": 28568802.0,
      "step": 16470
    },
    {
      "entropy": 5.577139949798584,
      "epoch": 1.281812876872204,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004838476812479968,
      "loss": 5.2143,
      "mean_token_accuracy": 0.18388948887586593,
      "num_tokens": 28578164.0,
      "step": 16475
    },
    {
      "entropy": 5.635396146774292,
      "epoch": 1.2822019062439214,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048383736802571774,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20787631571292878,
      "num_tokens": 28586061.0,
      "step": 16480
    },
    {
      "entropy": 5.547604417800903,
      "epoch": 1.282590935615639,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000483827051634664,
      "loss": 5.0404,
      "mean_token_accuracy": 0.192924864590168,
      "num_tokens": 28594185.0,
      "step": 16485
    },
    {
      "entropy": 5.546999168395996,
      "epoch": 1.2829799649873566,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048381673207499224,
      "loss": 5.0572,
      "mean_token_accuracy": 0.19981276541948317,
      "num_tokens": 28602427.0,
      "step": 16490
    },
    {
      "entropy": 5.534929466247559,
      "epoch": 1.283368994359074,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004838064093468588,
      "loss": 5.028,
      "mean_token_accuracy": 0.19716448038816453,
      "num_tokens": 28611608.0,
      "step": 16495
    },
    {
      "entropy": 5.548209953308105,
      "epoch": 1.2837580237307917,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004837960834504206,
      "loss": 5.0477,
      "mean_token_accuracy": 0.19497372210025787,
      "num_tokens": 28620142.0,
      "step": 16500
    },
    {
      "entropy": 5.604804468154907,
      "epoch": 1.2841470531025092,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004837857543858341,
      "loss": 5.1337,
      "mean_token_accuracy": 0.19307359606027602,
      "num_tokens": 28629401.0,
      "step": 16505
    },
    {
      "entropy": 5.589106655120849,
      "epoch": 1.2845360824742267,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004837754221532561,
      "loss": 5.0845,
      "mean_token_accuracy": 0.1930745080113411,
      "num_tokens": 28637934.0,
      "step": 16510
    },
    {
      "entropy": 5.586737060546875,
      "epoch": 1.2849251118459444,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048376508675284334,
      "loss": 5.0869,
      "mean_token_accuracy": 0.18979239016771315,
      "num_tokens": 28647145.0,
      "step": 16515
    },
    {
      "entropy": 5.608621883392334,
      "epoch": 1.2853141412176619,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048375474818475274,
      "loss": 5.1696,
      "mean_token_accuracy": 0.18448808491230012,
      "num_tokens": 28656146.0,
      "step": 16520
    },
    {
      "entropy": 5.584914875030518,
      "epoch": 1.2857031705893796,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048374440644914104,
      "loss": 5.0087,
      "mean_token_accuracy": 0.19568756818771363,
      "num_tokens": 28664476.0,
      "step": 16525
    },
    {
      "entropy": 5.606822443008423,
      "epoch": 1.286092199961097,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048373406154616523,
      "loss": 5.0378,
      "mean_token_accuracy": 0.19713133722543716,
      "num_tokens": 28672188.0,
      "step": 16530
    },
    {
      "entropy": 5.48529543876648,
      "epoch": 1.2864812293328147,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048372371347598226,
      "loss": 5.1033,
      "mean_token_accuracy": 0.19246701449155806,
      "num_tokens": 28680758.0,
      "step": 16535
    },
    {
      "entropy": 5.659683799743652,
      "epoch": 1.2868702587045322,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048371336223874923,
      "loss": 5.2103,
      "mean_token_accuracy": 0.17964619547128677,
      "num_tokens": 28690392.0,
      "step": 16540
    },
    {
      "entropy": 5.627428007125855,
      "epoch": 1.2872592880762497,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000483703007834623,
      "loss": 5.1011,
      "mean_token_accuracy": 0.1883941486477852,
      "num_tokens": 28699121.0,
      "step": 16545
    },
    {
      "entropy": 5.639080667495728,
      "epoch": 1.2876483174479674,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004836926502637609,
      "loss": 5.2099,
      "mean_token_accuracy": 0.18550428301095961,
      "num_tokens": 28707692.0,
      "step": 16550
    },
    {
      "entropy": 5.5936511039733885,
      "epoch": 1.2880373468196848,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048368228952632005,
      "loss": 5.1134,
      "mean_token_accuracy": 0.19203947633504867,
      "num_tokens": 28716630.0,
      "step": 16555
    },
    {
      "entropy": 5.559758853912354,
      "epoch": 1.2884263761914023,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048367192562245756,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20320075154304504,
      "num_tokens": 28725503.0,
      "step": 16560
    },
    {
      "entropy": 5.542258548736572,
      "epoch": 1.28881540556312,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048366155855233067,
      "loss": 5.1195,
      "mean_token_accuracy": 0.19535288363695144,
      "num_tokens": 28734578.0,
      "step": 16565
    },
    {
      "entropy": 5.595691537857055,
      "epoch": 1.2892044349348377,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004836511883160968,
      "loss": 5.1433,
      "mean_token_accuracy": 0.19096785932779312,
      "num_tokens": 28743380.0,
      "step": 16570
    },
    {
      "entropy": 5.602812099456787,
      "epoch": 1.2895934643065552,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004836408149139133,
      "loss": 5.107,
      "mean_token_accuracy": 0.19525352865457535,
      "num_tokens": 28752968.0,
      "step": 16575
    },
    {
      "entropy": 5.587865495681763,
      "epoch": 1.2899824936782727,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048363043834593743,
      "loss": 5.1271,
      "mean_token_accuracy": 0.1889372318983078,
      "num_tokens": 28761756.0,
      "step": 16580
    },
    {
      "entropy": 5.577973747253418,
      "epoch": 1.2903715230499904,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004836200586123268,
      "loss": 5.0909,
      "mean_token_accuracy": 0.1892427071928978,
      "num_tokens": 28771110.0,
      "step": 16585
    },
    {
      "entropy": 5.617769527435303,
      "epoch": 1.2907605524217078,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048360967571323875,
      "loss": 5.1663,
      "mean_token_accuracy": 0.19154311567544938,
      "num_tokens": 28780016.0,
      "step": 16590
    },
    {
      "entropy": 5.563744592666626,
      "epoch": 1.2911495817934253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048359928964883094,
      "loss": 5.0834,
      "mean_token_accuracy": 0.20017732232809066,
      "num_tokens": 28788150.0,
      "step": 16595
    },
    {
      "entropy": 5.584143352508545,
      "epoch": 1.291538611165143,
      "grad_norm": 1.15625,
      "learning_rate": 0.000483588900419261,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19261679351329802,
      "num_tokens": 28796564.0,
      "step": 16600
    },
    {
      "entropy": 5.6633295059204105,
      "epoch": 1.2919276405368605,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004835785080246864,
      "loss": 5.2654,
      "mean_token_accuracy": 0.1806255668401718,
      "num_tokens": 28804987.0,
      "step": 16605
    },
    {
      "entropy": 5.570775318145752,
      "epoch": 1.292316669908578,
      "grad_norm": 1.15625,
      "learning_rate": 0.000483568112465265,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19667039811611176,
      "num_tokens": 28813752.0,
      "step": 16610
    },
    {
      "entropy": 5.520872163772583,
      "epoch": 1.2927056992802957,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048355771374115436,
      "loss": 5.0088,
      "mean_token_accuracy": 0.19460532814264297,
      "num_tokens": 28822464.0,
      "step": 16615
    },
    {
      "entropy": 5.589781093597412,
      "epoch": 1.2930947286520134,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004835473118525125,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1967053607106209,
      "num_tokens": 28830997.0,
      "step": 16620
    },
    {
      "entropy": 5.551780891418457,
      "epoch": 1.2934837580237308,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004835369067994971,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20380938053131104,
      "num_tokens": 28839979.0,
      "step": 16625
    },
    {
      "entropy": 5.5718494892120365,
      "epoch": 1.2938727873954483,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000483526498582266,
      "loss": 5.0429,
      "mean_token_accuracy": 0.19419938176870347,
      "num_tokens": 28848394.0,
      "step": 16630
    },
    {
      "entropy": 5.670153284072876,
      "epoch": 1.294261816767166,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048351608720097715,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1793554037809372,
      "num_tokens": 28857804.0,
      "step": 16635
    },
    {
      "entropy": 5.5754327297210695,
      "epoch": 1.2946508461388835,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048350567265578867,
      "loss": 5.0187,
      "mean_token_accuracy": 0.1913873553276062,
      "num_tokens": 28866405.0,
      "step": 16640
    },
    {
      "entropy": 5.651254463195801,
      "epoch": 1.295039875510601,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004834952549468584,
      "loss": 5.1029,
      "mean_token_accuracy": 0.1922970697283745,
      "num_tokens": 28874751.0,
      "step": 16645
    },
    {
      "entropy": 5.602807998657227,
      "epoch": 1.2954289048823187,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004834848340743446,
      "loss": 5.2079,
      "mean_token_accuracy": 0.1860801100730896,
      "num_tokens": 28884013.0,
      "step": 16650
    },
    {
      "entropy": 5.570451402664185,
      "epoch": 1.2958179342540361,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004834744100384052,
      "loss": 5.0277,
      "mean_token_accuracy": 0.1930392399430275,
      "num_tokens": 28892175.0,
      "step": 16655
    },
    {
      "entropy": 5.623064851760864,
      "epoch": 1.2962069636257538,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004834639828391984,
      "loss": 5.0903,
      "mean_token_accuracy": 0.19090398401021957,
      "num_tokens": 28901238.0,
      "step": 16660
    },
    {
      "entropy": 5.6225049018859865,
      "epoch": 1.2965959929974713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048345355247688256,
      "loss": 5.0796,
      "mean_token_accuracy": 0.19284165203571318,
      "num_tokens": 28910368.0,
      "step": 16665
    },
    {
      "entropy": 5.596033573150635,
      "epoch": 1.296985022369189,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004834431189516158,
      "loss": 5.1572,
      "mean_token_accuracy": 0.18542687445878983,
      "num_tokens": 28918854.0,
      "step": 16670
    },
    {
      "entropy": 5.5833042621612545,
      "epoch": 1.2973740517409065,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004834326822635565,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19452373385429383,
      "num_tokens": 28927103.0,
      "step": 16675
    },
    {
      "entropy": 5.547755384445191,
      "epoch": 1.297763081112624,
      "grad_norm": 1.125,
      "learning_rate": 0.00048342224241286296,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19382714480161667,
      "num_tokens": 28936257.0,
      "step": 16680
    },
    {
      "entropy": 5.698548173904419,
      "epoch": 1.2981521104843416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004834117993996937,
      "loss": 5.1612,
      "mean_token_accuracy": 0.19172917753458024,
      "num_tokens": 28944991.0,
      "step": 16685
    },
    {
      "entropy": 5.643512916564942,
      "epoch": 1.2985411398560591,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004834013532242071,
      "loss": 5.1312,
      "mean_token_accuracy": 0.1850271463394165,
      "num_tokens": 28953781.0,
      "step": 16690
    },
    {
      "entropy": 5.585352993011474,
      "epoch": 1.2989301692277766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004833909038865616,
      "loss": 5.0497,
      "mean_token_accuracy": 0.1998119905591011,
      "num_tokens": 28961916.0,
      "step": 16695
    },
    {
      "entropy": 5.577362585067749,
      "epoch": 1.2993191985994943,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004833804513869159,
      "loss": 5.0658,
      "mean_token_accuracy": 0.18917641341686248,
      "num_tokens": 28970223.0,
      "step": 16700
    },
    {
      "entropy": 5.5597984790802,
      "epoch": 1.2997082279712118,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004833699957254284,
      "loss": 5.1123,
      "mean_token_accuracy": 0.19121940433979034,
      "num_tokens": 28979063.0,
      "step": 16705
    },
    {
      "entropy": 5.665278577804566,
      "epoch": 1.3000972573429295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000483359536902258,
      "loss": 5.2087,
      "mean_token_accuracy": 0.1856968432664871,
      "num_tokens": 28987229.0,
      "step": 16710
    },
    {
      "entropy": 5.652703332901001,
      "epoch": 1.300486286714647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004833490749175632,
      "loss": 5.1651,
      "mean_token_accuracy": 0.18514068871736528,
      "num_tokens": 28995889.0,
      "step": 16715
    },
    {
      "entropy": 5.618889427185058,
      "epoch": 1.3008753160863646,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048333860977150286,
      "loss": 5.1691,
      "mean_token_accuracy": 0.18045604079961777,
      "num_tokens": 29003962.0,
      "step": 16720
    },
    {
      "entropy": 5.559801197052002,
      "epoch": 1.3012643454580821,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048332814146423566,
      "loss": 5.0373,
      "mean_token_accuracy": 0.19711917340755464,
      "num_tokens": 29012652.0,
      "step": 16725
    },
    {
      "entropy": 5.684670877456665,
      "epoch": 1.3016533748297996,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004833176699959206,
      "loss": 5.1684,
      "mean_token_accuracy": 0.18112865090370178,
      "num_tokens": 29022445.0,
      "step": 16730
    },
    {
      "entropy": 5.586195468902588,
      "epoch": 1.3020424042015173,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048330719536671633,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19193349033594131,
      "num_tokens": 29030840.0,
      "step": 16735
    },
    {
      "entropy": 5.6061395645141605,
      "epoch": 1.3024314335732348,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000483296717576782,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19456223398447037,
      "num_tokens": 29039640.0,
      "step": 16740
    },
    {
      "entropy": 5.565610074996949,
      "epoch": 1.3028204629449522,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004832862366262765,
      "loss": 5.0756,
      "mean_token_accuracy": 0.1946895033121109,
      "num_tokens": 29047793.0,
      "step": 16745
    },
    {
      "entropy": 5.57900562286377,
      "epoch": 1.30320949231667,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048327575251535886,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19485248923301696,
      "num_tokens": 29056880.0,
      "step": 16750
    },
    {
      "entropy": 5.5813150882720945,
      "epoch": 1.3035985216883874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004832652652441883,
      "loss": 5.083,
      "mean_token_accuracy": 0.19047825038433075,
      "num_tokens": 29065093.0,
      "step": 16755
    },
    {
      "entropy": 5.572879457473755,
      "epoch": 1.303987551060105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048325477481292375,
      "loss": 5.1406,
      "mean_token_accuracy": 0.1924583837389946,
      "num_tokens": 29074389.0,
      "step": 16760
    },
    {
      "entropy": 5.484759950637818,
      "epoch": 1.3043765804318226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004832442812217244,
      "loss": 4.9948,
      "mean_token_accuracy": 0.20157041698694228,
      "num_tokens": 29082789.0,
      "step": 16765
    },
    {
      "entropy": 5.636271381378174,
      "epoch": 1.3047656098035403,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004832337844707496,
      "loss": 5.2517,
      "mean_token_accuracy": 0.17831240147352218,
      "num_tokens": 29092831.0,
      "step": 16770
    },
    {
      "entropy": 5.552855062484741,
      "epoch": 1.3051546391752578,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048322328456015855,
      "loss": 4.9503,
      "mean_token_accuracy": 0.19874482750892639,
      "num_tokens": 29101067.0,
      "step": 16775
    },
    {
      "entropy": 5.6161292552947994,
      "epoch": 1.3055436685469752,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048321278149011053,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1850826472043991,
      "num_tokens": 29110423.0,
      "step": 16780
    },
    {
      "entropy": 5.558938217163086,
      "epoch": 1.305932697918693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048320227526076504,
      "loss": 5.0174,
      "mean_token_accuracy": 0.19516593515872954,
      "num_tokens": 29119508.0,
      "step": 16785
    },
    {
      "entropy": 5.582490587234497,
      "epoch": 1.3063217272904104,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004831917658722814,
      "loss": 5.0999,
      "mean_token_accuracy": 0.19213947802782058,
      "num_tokens": 29127899.0,
      "step": 16790
    },
    {
      "entropy": 5.642197895050049,
      "epoch": 1.3067107566621279,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048318125332481903,
      "loss": 5.1305,
      "mean_token_accuracy": 0.19678532481193542,
      "num_tokens": 29137261.0,
      "step": 16795
    },
    {
      "entropy": 5.617922592163086,
      "epoch": 1.3070997860338456,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048317073761853764,
      "loss": 5.1949,
      "mean_token_accuracy": 0.18525552600622178,
      "num_tokens": 29145439.0,
      "step": 16800
    },
    {
      "entropy": 5.629019689559937,
      "epoch": 1.307488815405563,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004831602187535965,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1844330221414566,
      "num_tokens": 29153871.0,
      "step": 16805
    },
    {
      "entropy": 5.6042743682861325,
      "epoch": 1.3078778447772808,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004831496967301554,
      "loss": 5.0412,
      "mean_token_accuracy": 0.19668031185865403,
      "num_tokens": 29162343.0,
      "step": 16810
    },
    {
      "entropy": 5.561234951019287,
      "epoch": 1.3082668741489982,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048313917154837386,
      "loss": 5.1,
      "mean_token_accuracy": 0.19650373011827468,
      "num_tokens": 29170533.0,
      "step": 16815
    },
    {
      "entropy": 5.566080904006958,
      "epoch": 1.308655903520716,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004831286432084118,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19287993013858795,
      "num_tokens": 29179353.0,
      "step": 16820
    },
    {
      "entropy": 5.555334663391113,
      "epoch": 1.3090449328924334,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004831181117104289,
      "loss": 5.064,
      "mean_token_accuracy": 0.18950531780719757,
      "num_tokens": 29188747.0,
      "step": 16825
    },
    {
      "entropy": 5.564285230636597,
      "epoch": 1.3094339622641509,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048310757705458476,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1896413043141365,
      "num_tokens": 29197302.0,
      "step": 16830
    },
    {
      "entropy": 5.5804078578948975,
      "epoch": 1.3098229916358686,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048309703924103944,
      "loss": 5.0736,
      "mean_token_accuracy": 0.18789415508508683,
      "num_tokens": 29205667.0,
      "step": 16835
    },
    {
      "entropy": 5.638012790679932,
      "epoch": 1.310212021007586,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048308649826995283,
      "loss": 5.1542,
      "mean_token_accuracy": 0.1883809432387352,
      "num_tokens": 29214475.0,
      "step": 16840
    },
    {
      "entropy": 5.620544576644898,
      "epoch": 1.3106010503793035,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048307595414148475,
      "loss": 5.0598,
      "mean_token_accuracy": 0.19386268705129622,
      "num_tokens": 29222990.0,
      "step": 16845
    },
    {
      "entropy": 5.579739999771118,
      "epoch": 1.3109900797510212,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004830654068557952,
      "loss": 5.1557,
      "mean_token_accuracy": 0.19095881283283234,
      "num_tokens": 29232351.0,
      "step": 16850
    },
    {
      "entropy": 5.5729645729064945,
      "epoch": 1.3113791091227387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004830548564130444,
      "loss": 5.0503,
      "mean_token_accuracy": 0.19200800359249115,
      "num_tokens": 29240418.0,
      "step": 16855
    },
    {
      "entropy": 5.6306743144989015,
      "epoch": 1.3117681384944564,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048304430281339216,
      "loss": 5.1458,
      "mean_token_accuracy": 0.1861690640449524,
      "num_tokens": 29248975.0,
      "step": 16860
    },
    {
      "entropy": 5.59611668586731,
      "epoch": 1.3121571678661739,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004830337460569989,
      "loss": 5.1127,
      "mean_token_accuracy": 0.1880641832947731,
      "num_tokens": 29257866.0,
      "step": 16865
    },
    {
      "entropy": 5.507920122146606,
      "epoch": 1.3125461972378916,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004830231861440246,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21262308359146118,
      "num_tokens": 29266057.0,
      "step": 16870
    },
    {
      "entropy": 5.554885721206665,
      "epoch": 1.312935226609609,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004830126230746295,
      "loss": 5.0702,
      "mean_token_accuracy": 0.18819640427827836,
      "num_tokens": 29274752.0,
      "step": 16875
    },
    {
      "entropy": 5.674168825149536,
      "epoch": 1.3133242559813265,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048300205684897405,
      "loss": 5.1427,
      "mean_token_accuracy": 0.18988915234804155,
      "num_tokens": 29282781.0,
      "step": 16880
    },
    {
      "entropy": 5.550613212585449,
      "epoch": 1.3137132853530442,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004829914874672184,
      "loss": 5.0704,
      "mean_token_accuracy": 0.19598624259233474,
      "num_tokens": 29291196.0,
      "step": 16885
    },
    {
      "entropy": 5.6251153469085695,
      "epoch": 1.3141023147247617,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048298091492952297,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1862605482339859,
      "num_tokens": 29300168.0,
      "step": 16890
    },
    {
      "entropy": 5.569839191436768,
      "epoch": 1.3144913440964792,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004829703392360482,
      "loss": 5.0077,
      "mean_token_accuracy": 0.19489969611167907,
      "num_tokens": 29308468.0,
      "step": 16895
    },
    {
      "entropy": 5.525076103210449,
      "epoch": 1.3148803734681969,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048295976038695455,
      "loss": 5.0408,
      "mean_token_accuracy": 0.19172136336565018,
      "num_tokens": 29317096.0,
      "step": 16900
    },
    {
      "entropy": 5.659190559387207,
      "epoch": 1.3152694028399143,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004829491783824025,
      "loss": 5.1449,
      "mean_token_accuracy": 0.18676392436027528,
      "num_tokens": 29325840.0,
      "step": 16905
    },
    {
      "entropy": 5.559281301498413,
      "epoch": 1.315658432211632,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048293859322255276,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1942753791809082,
      "num_tokens": 29333874.0,
      "step": 16910
    },
    {
      "entropy": 5.555840349197387,
      "epoch": 1.3160474615833495,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004829280049075657,
      "loss": 5.0508,
      "mean_token_accuracy": 0.1904176354408264,
      "num_tokens": 29341937.0,
      "step": 16915
    },
    {
      "entropy": 5.64216980934143,
      "epoch": 1.3164364909550672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048291741343760216,
      "loss": 5.2777,
      "mean_token_accuracy": 0.1810837745666504,
      "num_tokens": 29350231.0,
      "step": 16920
    },
    {
      "entropy": 5.661997842788696,
      "epoch": 1.3168255203267847,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004829068188128229,
      "loss": 5.149,
      "mean_token_accuracy": 0.1888098046183586,
      "num_tokens": 29359036.0,
      "step": 16925
    },
    {
      "entropy": 5.649364423751831,
      "epoch": 1.3172145496985022,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004828962210333885,
      "loss": 5.032,
      "mean_token_accuracy": 0.19288186430931092,
      "num_tokens": 29368063.0,
      "step": 16930
    },
    {
      "entropy": 5.5886327743530275,
      "epoch": 1.3176035790702199,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004828856200994598,
      "loss": 5.0037,
      "mean_token_accuracy": 0.19591862708330154,
      "num_tokens": 29376653.0,
      "step": 16935
    },
    {
      "entropy": 5.49725604057312,
      "epoch": 1.3179926084419373,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004828750160111978,
      "loss": 5.0187,
      "mean_token_accuracy": 0.19977750182151793,
      "num_tokens": 29384978.0,
      "step": 16940
    },
    {
      "entropy": 5.6026331901550295,
      "epoch": 1.3183816378136548,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048286440876876325,
      "loss": 5.2,
      "mean_token_accuracy": 0.18370775282382965,
      "num_tokens": 29393927.0,
      "step": 16945
    },
    {
      "entropy": 5.7020350933074955,
      "epoch": 1.3187706671853725,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048285379837231714,
      "loss": 5.1681,
      "mean_token_accuracy": 0.19190392643213272,
      "num_tokens": 29402393.0,
      "step": 16950
    },
    {
      "entropy": 5.622466135025024,
      "epoch": 1.31915969655709,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004828431848220205,
      "loss": 5.1028,
      "mean_token_accuracy": 0.19281692057847977,
      "num_tokens": 29411011.0,
      "step": 16955
    },
    {
      "entropy": 5.622435808181763,
      "epoch": 1.3195487259288077,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004828325681180343,
      "loss": 5.1425,
      "mean_token_accuracy": 0.1904962554574013,
      "num_tokens": 29420095.0,
      "step": 16960
    },
    {
      "entropy": 5.564517688751221,
      "epoch": 1.3199377553005252,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004828219482605197,
      "loss": 5.1252,
      "mean_token_accuracy": 0.19043058604002,
      "num_tokens": 29429274.0,
      "step": 16965
    },
    {
      "entropy": 5.618411588668823,
      "epoch": 1.3203267846722428,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048281132524963786,
      "loss": 5.0959,
      "mean_token_accuracy": 0.18482347279787065,
      "num_tokens": 29437572.0,
      "step": 16970
    },
    {
      "entropy": 5.604471063613891,
      "epoch": 1.3207158140439603,
      "grad_norm": 1.125,
      "learning_rate": 0.0004828006990855499,
      "loss": 5.1086,
      "mean_token_accuracy": 0.19881976693868636,
      "num_tokens": 29446294.0,
      "step": 16975
    },
    {
      "entropy": 5.5644371509552,
      "epoch": 1.3211048434156778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048279006976841704,
      "loss": 5.1119,
      "mean_token_accuracy": 0.1945158302783966,
      "num_tokens": 29454710.0,
      "step": 16980
    },
    {
      "entropy": 5.613303375244141,
      "epoch": 1.3214938727873955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004827794372984007,
      "loss": 5.1217,
      "mean_token_accuracy": 0.19107856154441832,
      "num_tokens": 29463067.0,
      "step": 16985
    },
    {
      "entropy": 5.63302845954895,
      "epoch": 1.321882902159113,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000482768801675662,
      "loss": 5.1129,
      "mean_token_accuracy": 0.18330609798431396,
      "num_tokens": 29471636.0,
      "step": 16990
    },
    {
      "entropy": 5.623661947250366,
      "epoch": 1.3222719315308304,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004827581629003625,
      "loss": 5.1226,
      "mean_token_accuracy": 0.19127574563026428,
      "num_tokens": 29480772.0,
      "step": 16995
    },
    {
      "entropy": 5.554599618911743,
      "epoch": 1.3226609609025481,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048274752097266356,
      "loss": 5.0314,
      "mean_token_accuracy": 0.196151240170002,
      "num_tokens": 29489113.0,
      "step": 17000
    },
    {
      "entropy": 5.620364141464234,
      "epoch": 1.3230499902742658,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004827368758927266,
      "loss": 5.1046,
      "mean_token_accuracy": 0.1901696190237999,
      "num_tokens": 29497721.0,
      "step": 17005
    },
    {
      "entropy": 5.58538122177124,
      "epoch": 1.3234390196459833,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004827262276607132,
      "loss": 5.0818,
      "mean_token_accuracy": 0.1918430969119072,
      "num_tokens": 29506452.0,
      "step": 17010
    },
    {
      "entropy": 5.566258573532105,
      "epoch": 1.3238280490177008,
      "grad_norm": 1.109375,
      "learning_rate": 0.000482715576276785,
      "loss": 5.0506,
      "mean_token_accuracy": 0.19990090429782867,
      "num_tokens": 29514635.0,
      "step": 17015
    },
    {
      "entropy": 5.562077331542969,
      "epoch": 1.3242170783894185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004827049217411035,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19293052554130555,
      "num_tokens": 29523929.0,
      "step": 17020
    },
    {
      "entropy": 5.593743801116943,
      "epoch": 1.324606107761136,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048269426405383043,
      "loss": 5.0657,
      "mean_token_accuracy": 0.19113940447568895,
      "num_tokens": 29533125.0,
      "step": 17025
    },
    {
      "entropy": 5.62114462852478,
      "epoch": 1.3249951371328534,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048268360321512745,
      "loss": 5.079,
      "mean_token_accuracy": 0.2026069775223732,
      "num_tokens": 29541604.0,
      "step": 17030
    },
    {
      "entropy": 5.602646541595459,
      "epoch": 1.3253841665045711,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004826729392251564,
      "loss": 5.0826,
      "mean_token_accuracy": 0.19362288415431977,
      "num_tokens": 29550100.0,
      "step": 17035
    },
    {
      "entropy": 5.592345952987671,
      "epoch": 1.3257731958762886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000482662272084079,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1944161370396614,
      "num_tokens": 29559490.0,
      "step": 17040
    },
    {
      "entropy": 5.660453033447266,
      "epoch": 1.326162225248006,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004826516017920571,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1833604544401169,
      "num_tokens": 29568734.0,
      "step": 17045
    },
    {
      "entropy": 5.590445327758789,
      "epoch": 1.3265512546197238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004826409283492528,
      "loss": 5.1173,
      "mean_token_accuracy": 0.1985136851668358,
      "num_tokens": 29577339.0,
      "step": 17050
    },
    {
      "entropy": 5.602376222610474,
      "epoch": 1.3269402839914415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048263025175582783,
      "loss": 5.2182,
      "mean_token_accuracy": 0.18692563474178314,
      "num_tokens": 29585561.0,
      "step": 17055
    },
    {
      "entropy": 5.600705575942993,
      "epoch": 1.327329313363159,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004826195720119442,
      "loss": 5.1657,
      "mean_token_accuracy": 0.19134705364704133,
      "num_tokens": 29594663.0,
      "step": 17060
    },
    {
      "entropy": 5.646587657928467,
      "epoch": 1.3277183427348764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004826088891177641,
      "loss": 5.1238,
      "mean_token_accuracy": 0.19649520367383957,
      "num_tokens": 29603537.0,
      "step": 17065
    },
    {
      "entropy": 5.647028112411499,
      "epoch": 1.3281073721065941,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048259820307344954,
      "loss": 5.1301,
      "mean_token_accuracy": 0.19342322498559952,
      "num_tokens": 29612234.0,
      "step": 17070
    },
    {
      "entropy": 5.5453869819641115,
      "epoch": 1.3284964014783116,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048258751387916267,
      "loss": 4.9578,
      "mean_token_accuracy": 0.19535627961158752,
      "num_tokens": 29621435.0,
      "step": 17075
    },
    {
      "entropy": 5.712816333770752,
      "epoch": 1.328885430850029,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048257682153506564,
      "loss": 5.2364,
      "mean_token_accuracy": 0.18353254348039627,
      "num_tokens": 29630266.0,
      "step": 17080
    },
    {
      "entropy": 5.514873886108399,
      "epoch": 1.3292744602217468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004825661260413208,
      "loss": 5.0222,
      "mean_token_accuracy": 0.19335971176624298,
      "num_tokens": 29639118.0,
      "step": 17085
    },
    {
      "entropy": 5.5482714653015135,
      "epoch": 1.3296634895934643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048255542739809035,
      "loss": 5.1109,
      "mean_token_accuracy": 0.2015178069472313,
      "num_tokens": 29647616.0,
      "step": 17090
    },
    {
      "entropy": 5.610386610031128,
      "epoch": 1.330052518965182,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048254472560553665,
      "loss": 5.0636,
      "mean_token_accuracy": 0.2000719740986824,
      "num_tokens": 29656264.0,
      "step": 17095
    },
    {
      "entropy": 5.631341028213501,
      "epoch": 1.3304415483368994,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048253402066382207,
      "loss": 5.0728,
      "mean_token_accuracy": 0.19912388026714326,
      "num_tokens": 29664452.0,
      "step": 17100
    },
    {
      "entropy": 5.545564413070679,
      "epoch": 1.3308305777086171,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004825233125731091,
      "loss": 5.0289,
      "mean_token_accuracy": 0.19733412563800812,
      "num_tokens": 29673456.0,
      "step": 17105
    },
    {
      "entropy": 5.585660743713379,
      "epoch": 1.3312196070803346,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048251260133356014,
      "loss": 5.1451,
      "mean_token_accuracy": 0.18881290704011916,
      "num_tokens": 29682232.0,
      "step": 17110
    },
    {
      "entropy": 5.5996520042419435,
      "epoch": 1.331608636452052,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048250188694533774,
      "loss": 5.0695,
      "mean_token_accuracy": 0.19354334622621536,
      "num_tokens": 29690421.0,
      "step": 17115
    },
    {
      "entropy": 5.5877831935882565,
      "epoch": 1.3319976658237698,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004824911694086044,
      "loss": 5.0611,
      "mean_token_accuracy": 0.1953008159995079,
      "num_tokens": 29698844.0,
      "step": 17120
    },
    {
      "entropy": 5.593536376953125,
      "epoch": 1.3323866951954872,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000482480448723523,
      "loss": 5.0567,
      "mean_token_accuracy": 0.18919510990381241,
      "num_tokens": 29706738.0,
      "step": 17125
    },
    {
      "entropy": 5.62646746635437,
      "epoch": 1.3327757245672047,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048246972489025586,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1917755126953125,
      "num_tokens": 29715812.0,
      "step": 17130
    },
    {
      "entropy": 5.5427632331848145,
      "epoch": 1.3331647539389224,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048245899790896597,
      "loss": 5.0494,
      "mean_token_accuracy": 0.1940422460436821,
      "num_tokens": 29724488.0,
      "step": 17135
    },
    {
      "entropy": 5.5217279434204105,
      "epoch": 1.33355378331064,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004824482677798159,
      "loss": 5.0218,
      "mean_token_accuracy": 0.19493692368268967,
      "num_tokens": 29733565.0,
      "step": 17140
    },
    {
      "entropy": 5.577882528305054,
      "epoch": 1.3339428126823576,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048243753450296863,
      "loss": 5.1035,
      "mean_token_accuracy": 0.19251312613487243,
      "num_tokens": 29742437.0,
      "step": 17145
    },
    {
      "entropy": 5.571154975891114,
      "epoch": 1.334331842054075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048242679807858693,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19604819864034653,
      "num_tokens": 29751636.0,
      "step": 17150
    },
    {
      "entropy": 5.596826648712158,
      "epoch": 1.3347208714257928,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048241605850683365,
      "loss": 5.1101,
      "mean_token_accuracy": 0.1892961949110031,
      "num_tokens": 29760199.0,
      "step": 17155
    },
    {
      "entropy": 5.6211920261383055,
      "epoch": 1.3351099007975102,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048240531578787197,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19168538749217987,
      "num_tokens": 29769169.0,
      "step": 17160
    },
    {
      "entropy": 5.56289849281311,
      "epoch": 1.3354989301692277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004823945699218647,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20089684277772904,
      "num_tokens": 29777218.0,
      "step": 17165
    },
    {
      "entropy": 5.579970598220825,
      "epoch": 1.3358879595409454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004823838209089749,
      "loss": 5.074,
      "mean_token_accuracy": 0.19264025837183,
      "num_tokens": 29786221.0,
      "step": 17170
    },
    {
      "entropy": 5.617652034759521,
      "epoch": 1.3362769889126629,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048237306874936565,
      "loss": 5.093,
      "mean_token_accuracy": 0.19238927811384202,
      "num_tokens": 29794659.0,
      "step": 17175
    },
    {
      "entropy": 5.662933683395385,
      "epoch": 1.3366660182843804,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004823623134432001,
      "loss": 5.0475,
      "mean_token_accuracy": 0.19825076162815095,
      "num_tokens": 29804237.0,
      "step": 17180
    },
    {
      "entropy": 5.672011137008667,
      "epoch": 1.337055047656098,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048235155499064166,
      "loss": 5.2025,
      "mean_token_accuracy": 0.18443514108657838,
      "num_tokens": 29813521.0,
      "step": 17185
    },
    {
      "entropy": 5.644973850250244,
      "epoch": 1.3374440770278155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048234079339185335,
      "loss": 5.2638,
      "mean_token_accuracy": 0.1848655104637146,
      "num_tokens": 29822123.0,
      "step": 17190
    },
    {
      "entropy": 5.590908145904541,
      "epoch": 1.3378331063995332,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048233002864699856,
      "loss": 5.1739,
      "mean_token_accuracy": 0.18695790320634842,
      "num_tokens": 29830557.0,
      "step": 17195
    },
    {
      "entropy": 5.588746118545532,
      "epoch": 1.3382221357712507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004823192607562405,
      "loss": 5.0811,
      "mean_token_accuracy": 0.19485005587339402,
      "num_tokens": 29839566.0,
      "step": 17200
    },
    {
      "entropy": 5.638722562789917,
      "epoch": 1.3386111651429684,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048230848971974265,
      "loss": 5.0583,
      "mean_token_accuracy": 0.19577376842498778,
      "num_tokens": 29847808.0,
      "step": 17205
    },
    {
      "entropy": 5.5864400386810305,
      "epoch": 1.3390001945146859,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004822977155376684,
      "loss": 5.0209,
      "mean_token_accuracy": 0.1998085707426071,
      "num_tokens": 29855974.0,
      "step": 17210
    },
    {
      "entropy": 5.659725141525269,
      "epoch": 1.3393892238864034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004822869382101814,
      "loss": 5.1376,
      "mean_token_accuracy": 0.19154758602380753,
      "num_tokens": 29864910.0,
      "step": 17215
    },
    {
      "entropy": 5.625275802612305,
      "epoch": 1.339778253258121,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004822761577374449,
      "loss": 5.2314,
      "mean_token_accuracy": 0.19052161425352096,
      "num_tokens": 29873668.0,
      "step": 17220
    },
    {
      "entropy": 5.622101449966431,
      "epoch": 1.3401672826298385,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004822653741196227,
      "loss": 5.15,
      "mean_token_accuracy": 0.1858844131231308,
      "num_tokens": 29882099.0,
      "step": 17225
    },
    {
      "entropy": 5.602360582351684,
      "epoch": 1.340556312001556,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004822545873568783,
      "loss": 5.051,
      "mean_token_accuracy": 0.19473132789134978,
      "num_tokens": 29891022.0,
      "step": 17230
    },
    {
      "entropy": 5.601638269424439,
      "epoch": 1.3409453413732737,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048224379744937545,
      "loss": 5.0696,
      "mean_token_accuracy": 0.192145274579525,
      "num_tokens": 29899270.0,
      "step": 17235
    },
    {
      "entropy": 5.676001930236817,
      "epoch": 1.3413343707449912,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048223300439727783,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18344607055187226,
      "num_tokens": 29908183.0,
      "step": 17240
    },
    {
      "entropy": 5.615677452087402,
      "epoch": 1.3417234001167089,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048222220820074926,
      "loss": 5.1269,
      "mean_token_accuracy": 0.18924863636493683,
      "num_tokens": 29917148.0,
      "step": 17245
    },
    {
      "entropy": 5.5948240756988525,
      "epoch": 1.3421124294884264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048221140885995346,
      "loss": 5.0211,
      "mean_token_accuracy": 0.1978804886341095,
      "num_tokens": 29926251.0,
      "step": 17250
    },
    {
      "entropy": 5.6015252590179445,
      "epoch": 1.342501458860144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004822006063750543,
      "loss": 5.1231,
      "mean_token_accuracy": 0.18970627337694168,
      "num_tokens": 29935000.0,
      "step": 17255
    },
    {
      "entropy": 5.5656873226165775,
      "epoch": 1.3428904882318615,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048218980074621575,
      "loss": 5.001,
      "mean_token_accuracy": 0.2058898463845253,
      "num_tokens": 29943046.0,
      "step": 17260
    },
    {
      "entropy": 5.611284923553467,
      "epoch": 1.343279517603579,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048217899197360183,
      "loss": 5.1095,
      "mean_token_accuracy": 0.19352975338697434,
      "num_tokens": 29952254.0,
      "step": 17265
    },
    {
      "entropy": 5.587187814712524,
      "epoch": 1.3436685469752967,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004821681800573764,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19497756212949752,
      "num_tokens": 29961458.0,
      "step": 17270
    },
    {
      "entropy": 5.6207112789154055,
      "epoch": 1.3440575763470142,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004821573649977036,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20127401053905486,
      "num_tokens": 29969865.0,
      "step": 17275
    },
    {
      "entropy": 5.596780729293823,
      "epoch": 1.3444466057187316,
      "grad_norm": 1.125,
      "learning_rate": 0.00048214654679474753,
      "loss": 5.0866,
      "mean_token_accuracy": 0.18768148273229598,
      "num_tokens": 29978260.0,
      "step": 17280
    },
    {
      "entropy": 5.654696321487426,
      "epoch": 1.3448356350904493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048213572544867224,
      "loss": 5.2173,
      "mean_token_accuracy": 0.1866936832666397,
      "num_tokens": 29987528.0,
      "step": 17285
    },
    {
      "entropy": 5.581232261657715,
      "epoch": 1.3452246644621668,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048212490095964213,
      "loss": 5.081,
      "mean_token_accuracy": 0.18714954257011412,
      "num_tokens": 29995991.0,
      "step": 17290
    },
    {
      "entropy": 5.6148152351379395,
      "epoch": 1.3456136938338845,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004821140733278212,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20062678754329683,
      "num_tokens": 30004569.0,
      "step": 17295
    },
    {
      "entropy": 5.597702741622925,
      "epoch": 1.346002723205602,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004821032425533739,
      "loss": 5.1196,
      "mean_token_accuracy": 0.191415037214756,
      "num_tokens": 30014021.0,
      "step": 17300
    },
    {
      "entropy": 5.526555347442627,
      "epoch": 1.3463917525773197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004820924086364646,
      "loss": 4.946,
      "mean_token_accuracy": 0.19913501143455506,
      "num_tokens": 30021803.0,
      "step": 17305
    },
    {
      "entropy": 5.569703817367554,
      "epoch": 1.3467807819490372,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048208157157725756,
      "loss": 5.0964,
      "mean_token_accuracy": 0.1948235362768173,
      "num_tokens": 30030707.0,
      "step": 17310
    },
    {
      "entropy": 5.6308599472045895,
      "epoch": 1.3471698113207546,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048207073137591726,
      "loss": 5.1249,
      "mean_token_accuracy": 0.18737078458070755,
      "num_tokens": 30040235.0,
      "step": 17315
    },
    {
      "entropy": 5.676722621917724,
      "epoch": 1.3475588406924723,
      "grad_norm": 1.125,
      "learning_rate": 0.00048205988803260824,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19530408680438996,
      "num_tokens": 30049361.0,
      "step": 17320
    },
    {
      "entropy": 5.600044536590576,
      "epoch": 1.3479478700641898,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048204904154749496,
      "loss": 5.1371,
      "mean_token_accuracy": 0.18390728086233138,
      "num_tokens": 30058896.0,
      "step": 17325
    },
    {
      "entropy": 5.60433874130249,
      "epoch": 1.3483368994359073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048203819192074206,
      "loss": 5.165,
      "mean_token_accuracy": 0.18034772723913192,
      "num_tokens": 30067206.0,
      "step": 17330
    },
    {
      "entropy": 5.605970096588135,
      "epoch": 1.348725928807625,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048202733915251407,
      "loss": 5.0405,
      "mean_token_accuracy": 0.19634858518838882,
      "num_tokens": 30076286.0,
      "step": 17335
    },
    {
      "entropy": 5.516799259185791,
      "epoch": 1.3491149581793425,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048201648324297573,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2011720910668373,
      "num_tokens": 30084681.0,
      "step": 17340
    },
    {
      "entropy": 5.5704261302948,
      "epoch": 1.3495039875510602,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048200562419229185,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1911061704158783,
      "num_tokens": 30093293.0,
      "step": 17345
    },
    {
      "entropy": 5.520547151565552,
      "epoch": 1.3498930169227776,
      "grad_norm": 1.03125,
      "learning_rate": 0.000481994762000627,
      "loss": 4.9399,
      "mean_token_accuracy": 0.20670391470193863,
      "num_tokens": 30101593.0,
      "step": 17350
    },
    {
      "entropy": 5.593836545944214,
      "epoch": 1.3502820462944953,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048198389666814616,
      "loss": 5.093,
      "mean_token_accuracy": 0.19063035845756532,
      "num_tokens": 30110328.0,
      "step": 17355
    },
    {
      "entropy": 5.555759048461914,
      "epoch": 1.3506710756662128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048197302819501416,
      "loss": 5.0687,
      "mean_token_accuracy": 0.19296697080135344,
      "num_tokens": 30118721.0,
      "step": 17360
    },
    {
      "entropy": 5.532725811004639,
      "epoch": 1.3510601050379303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048196215658139586,
      "loss": 5.0023,
      "mean_token_accuracy": 0.19072745591402054,
      "num_tokens": 30127238.0,
      "step": 17365
    },
    {
      "entropy": 5.5552952766418455,
      "epoch": 1.351449134409648,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004819512818274562,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20025107115507126,
      "num_tokens": 30135049.0,
      "step": 17370
    },
    {
      "entropy": 5.4552083015441895,
      "epoch": 1.3518381637813655,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004819404039333603,
      "loss": 5.0644,
      "mean_token_accuracy": 0.19416424334049226,
      "num_tokens": 30143991.0,
      "step": 17375
    },
    {
      "entropy": 5.584801864624024,
      "epoch": 1.352227193153083,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004819295228992731,
      "loss": 5.2155,
      "mean_token_accuracy": 0.18293273597955703,
      "num_tokens": 30153968.0,
      "step": 17380
    },
    {
      "entropy": 5.665489959716797,
      "epoch": 1.3526162225248006,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048191863872535984,
      "loss": 5.098,
      "mean_token_accuracy": 0.1899640828371048,
      "num_tokens": 30162927.0,
      "step": 17385
    },
    {
      "entropy": 5.527276849746704,
      "epoch": 1.3530052518965183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004819077514117855,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1998799502849579,
      "num_tokens": 30170986.0,
      "step": 17390
    },
    {
      "entropy": 5.527010536193847,
      "epoch": 1.3533942812682358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048189686095871545,
      "loss": 5.0289,
      "mean_token_accuracy": 0.1994290128350258,
      "num_tokens": 30179167.0,
      "step": 17395
    },
    {
      "entropy": 5.512463665008545,
      "epoch": 1.3537833106399533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048188596736631475,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1981296494603157,
      "num_tokens": 30187710.0,
      "step": 17400
    },
    {
      "entropy": 5.609972524642944,
      "epoch": 1.354172340011671,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048187507063474885,
      "loss": 5.1329,
      "mean_token_accuracy": 0.19199165999889373,
      "num_tokens": 30196502.0,
      "step": 17405
    },
    {
      "entropy": 5.523221063613891,
      "epoch": 1.3545613693833884,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048186417076418294,
      "loss": 5.012,
      "mean_token_accuracy": 0.19531462490558624,
      "num_tokens": 30205070.0,
      "step": 17410
    },
    {
      "entropy": 5.617769575119018,
      "epoch": 1.354950398755106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048185326775478256,
      "loss": 5.0751,
      "mean_token_accuracy": 0.1883956030011177,
      "num_tokens": 30213554.0,
      "step": 17415
    },
    {
      "entropy": 5.537787628173828,
      "epoch": 1.3553394281268236,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048184236160671306,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20149375647306442,
      "num_tokens": 30222687.0,
      "step": 17420
    },
    {
      "entropy": 5.56029691696167,
      "epoch": 1.355728457498541,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048183145232013995,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1900578796863556,
      "num_tokens": 30231086.0,
      "step": 17425
    },
    {
      "entropy": 5.534148931503296,
      "epoch": 1.3561174868702586,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048182053989522883,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2029384434223175,
      "num_tokens": 30240355.0,
      "step": 17430
    },
    {
      "entropy": 5.652457332611084,
      "epoch": 1.3565065162419763,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048180962433214515,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19059297889471055,
      "num_tokens": 30249586.0,
      "step": 17435
    },
    {
      "entropy": 5.630220985412597,
      "epoch": 1.356895545613694,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048179870563105453,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19273159205913543,
      "num_tokens": 30257666.0,
      "step": 17440
    },
    {
      "entropy": 5.612457323074341,
      "epoch": 1.3572845749854114,
      "grad_norm": 1.125,
      "learning_rate": 0.0004817877837921228,
      "loss": 5.2461,
      "mean_token_accuracy": 0.19070723950862883,
      "num_tokens": 30266846.0,
      "step": 17445
    },
    {
      "entropy": 5.5119424819946286,
      "epoch": 1.357673604357129,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004817768588155155,
      "loss": 4.9929,
      "mean_token_accuracy": 0.1987408861517906,
      "num_tokens": 30275587.0,
      "step": 17450
    },
    {
      "entropy": 5.602882385253906,
      "epoch": 1.3580626337288466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048176593070139865,
      "loss": 5.0657,
      "mean_token_accuracy": 0.2000027909874916,
      "num_tokens": 30283948.0,
      "step": 17455
    },
    {
      "entropy": 5.697951412200927,
      "epoch": 1.358451663100564,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048175499944993767,
      "loss": 5.2267,
      "mean_token_accuracy": 0.1879907727241516,
      "num_tokens": 30292302.0,
      "step": 17460
    },
    {
      "entropy": 5.638585615158081,
      "epoch": 1.3588406924722816,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048174406506129884,
      "loss": 5.1956,
      "mean_token_accuracy": 0.18729890286922454,
      "num_tokens": 30301414.0,
      "step": 17465
    },
    {
      "entropy": 5.64165186882019,
      "epoch": 1.3592297218439993,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048173312753564787,
      "loss": 5.0838,
      "mean_token_accuracy": 0.207299642264843,
      "num_tokens": 30311045.0,
      "step": 17470
    },
    {
      "entropy": 5.632576274871826,
      "epoch": 1.3596187512157167,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004817221868731506,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19632697552442552,
      "num_tokens": 30318552.0,
      "step": 17475
    },
    {
      "entropy": 5.505974245071411,
      "epoch": 1.3600077805874344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048171124307397334,
      "loss": 5.1116,
      "mean_token_accuracy": 0.19565628916025163,
      "num_tokens": 30327583.0,
      "step": 17480
    },
    {
      "entropy": 5.578305959701538,
      "epoch": 1.360396809959152,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004817002961382819,
      "loss": 5.0746,
      "mean_token_accuracy": 0.18917604833841323,
      "num_tokens": 30336339.0,
      "step": 17485
    },
    {
      "entropy": 5.672673845291138,
      "epoch": 1.3607858393308696,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048168934606624255,
      "loss": 5.1376,
      "mean_token_accuracy": 0.1942661687731743,
      "num_tokens": 30344878.0,
      "step": 17490
    },
    {
      "entropy": 5.612644958496094,
      "epoch": 1.361174868702587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048167839285802116,
      "loss": 5.0987,
      "mean_token_accuracy": 0.19125708043575287,
      "num_tokens": 30353071.0,
      "step": 17495
    },
    {
      "entropy": 5.557382202148437,
      "epoch": 1.3615638980743046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004816674365137843,
      "loss": 5.076,
      "mean_token_accuracy": 0.19334767162799835,
      "num_tokens": 30361356.0,
      "step": 17500
    },
    {
      "entropy": 5.605579805374146,
      "epoch": 1.3619529274460223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000481656477033698,
      "loss": 5.1142,
      "mean_token_accuracy": 0.18398269563913344,
      "num_tokens": 30370917.0,
      "step": 17505
    },
    {
      "entropy": 5.602678632736206,
      "epoch": 1.3623419568177397,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004816455144179286,
      "loss": 5.1233,
      "mean_token_accuracy": 0.1865637332201004,
      "num_tokens": 30380304.0,
      "step": 17510
    },
    {
      "entropy": 5.640541887283325,
      "epoch": 1.3627309861894572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004816345486666424,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19976193755865096,
      "num_tokens": 30388565.0,
      "step": 17515
    },
    {
      "entropy": 5.6112456798553465,
      "epoch": 1.363120015561175,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004816235797800058,
      "loss": 5.209,
      "mean_token_accuracy": 0.18759981393814087,
      "num_tokens": 30397220.0,
      "step": 17520
    },
    {
      "entropy": 5.581954145431519,
      "epoch": 1.3635090449328924,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004816126077581852,
      "loss": 5.061,
      "mean_token_accuracy": 0.19477532356977462,
      "num_tokens": 30405235.0,
      "step": 17525
    },
    {
      "entropy": 5.586590576171875,
      "epoch": 1.36389807430461,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048160163260134724,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19074337035417557,
      "num_tokens": 30414218.0,
      "step": 17530
    },
    {
      "entropy": 5.58496503829956,
      "epoch": 1.3642871036763276,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004815906543096583,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19168416112661363,
      "num_tokens": 30422665.0,
      "step": 17535
    },
    {
      "entropy": 5.627125549316406,
      "epoch": 1.3646761330480452,
      "grad_norm": 1.03125,
      "learning_rate": 0.000481579672883285,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19659970998764037,
      "num_tokens": 30431868.0,
      "step": 17540
    },
    {
      "entropy": 5.620336675643921,
      "epoch": 1.3650651624197627,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048156868832239396,
      "loss": 5.1107,
      "mean_token_accuracy": 0.192998768389225,
      "num_tokens": 30440548.0,
      "step": 17545
    },
    {
      "entropy": 5.668788290023803,
      "epoch": 1.3654541917914802,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004815577006271519,
      "loss": 5.129,
      "mean_token_accuracy": 0.18544111549854278,
      "num_tokens": 30448959.0,
      "step": 17550
    },
    {
      "entropy": 5.590928459167481,
      "epoch": 1.365843221163198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048154670979772555,
      "loss": 5.0257,
      "mean_token_accuracy": 0.19717933237552643,
      "num_tokens": 30457193.0,
      "step": 17555
    },
    {
      "entropy": 5.687171506881714,
      "epoch": 1.3662322505349154,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004815357158342815,
      "loss": 5.0866,
      "mean_token_accuracy": 0.19243291914463043,
      "num_tokens": 30466253.0,
      "step": 17560
    },
    {
      "entropy": 5.588253831863403,
      "epoch": 1.3666212799066328,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004815247187369868,
      "loss": 5.0236,
      "mean_token_accuracy": 0.1960255667567253,
      "num_tokens": 30474675.0,
      "step": 17565
    },
    {
      "entropy": 5.563898229598999,
      "epoch": 1.3670103092783505,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048151371850600814,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19621526300907136,
      "num_tokens": 30483031.0,
      "step": 17570
    },
    {
      "entropy": 5.51670093536377,
      "epoch": 1.367399338650068,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048150271514151255,
      "loss": 5.0678,
      "mean_token_accuracy": 0.19688308537006377,
      "num_tokens": 30491331.0,
      "step": 17575
    },
    {
      "entropy": 5.5813130855560305,
      "epoch": 1.3677883680217857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048149170864366693,
      "loss": 5.0848,
      "mean_token_accuracy": 0.1881057322025299,
      "num_tokens": 30499649.0,
      "step": 17580
    },
    {
      "entropy": 5.590866804122925,
      "epoch": 1.3681773973935032,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048148069901263836,
      "loss": 4.9617,
      "mean_token_accuracy": 0.19990209937095643,
      "num_tokens": 30508922.0,
      "step": 17585
    },
    {
      "entropy": 5.588878011703491,
      "epoch": 1.368566426765221,
      "grad_norm": 1.125,
      "learning_rate": 0.00048146968624859375,
      "loss": 5.0446,
      "mean_token_accuracy": 0.1936262220144272,
      "num_tokens": 30517745.0,
      "step": 17590
    },
    {
      "entropy": 5.589673137664795,
      "epoch": 1.3689554561369384,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004814586703517003,
      "loss": 5.1782,
      "mean_token_accuracy": 0.18399705737829208,
      "num_tokens": 30526597.0,
      "step": 17595
    },
    {
      "entropy": 5.665196180343628,
      "epoch": 1.3693444855086558,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004814476513221251,
      "loss": 5.0707,
      "mean_token_accuracy": 0.19616006016731263,
      "num_tokens": 30533969.0,
      "step": 17600
    },
    {
      "entropy": 5.626406860351563,
      "epoch": 1.3697335148803735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048143662916003545,
      "loss": 5.1489,
      "mean_token_accuracy": 0.1897202730178833,
      "num_tokens": 30543277.0,
      "step": 17605
    },
    {
      "entropy": 5.587683248519897,
      "epoch": 1.370122544252091,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004814256038655985,
      "loss": 5.1039,
      "mean_token_accuracy": 0.1873035430908203,
      "num_tokens": 30551889.0,
      "step": 17610
    },
    {
      "entropy": 5.607007694244385,
      "epoch": 1.3705115736238085,
      "grad_norm": 1.125,
      "learning_rate": 0.0004814145754389816,
      "loss": 5.0918,
      "mean_token_accuracy": 0.19290568977594375,
      "num_tokens": 30560926.0,
      "step": 17615
    },
    {
      "entropy": 5.72218165397644,
      "epoch": 1.3709006029955262,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048140354388035204,
      "loss": 5.1582,
      "mean_token_accuracy": 0.18969112783670425,
      "num_tokens": 30570102.0,
      "step": 17620
    },
    {
      "entropy": 5.605351448059082,
      "epoch": 1.3712896323672437,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004813925091898772,
      "loss": 5.1089,
      "mean_token_accuracy": 0.19315589517354964,
      "num_tokens": 30578585.0,
      "step": 17625
    },
    {
      "entropy": 5.585934066772461,
      "epoch": 1.3716786617389614,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004813814713677246,
      "loss": 5.1085,
      "mean_token_accuracy": 0.19218945056200026,
      "num_tokens": 30587594.0,
      "step": 17630
    },
    {
      "entropy": 5.681205320358276,
      "epoch": 1.3720676911106788,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004813704304140616,
      "loss": 5.1209,
      "mean_token_accuracy": 0.19022925049066544,
      "num_tokens": 30596234.0,
      "step": 17635
    },
    {
      "entropy": 5.657306814193726,
      "epoch": 1.3724567204823965,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004813593863290559,
      "loss": 5.1625,
      "mean_token_accuracy": 0.18896152824163437,
      "num_tokens": 30605822.0,
      "step": 17640
    },
    {
      "entropy": 5.5634666919708256,
      "epoch": 1.372845749854114,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004813483391128748,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19636477380990983,
      "num_tokens": 30614154.0,
      "step": 17645
    },
    {
      "entropy": 5.646077394485474,
      "epoch": 1.3732347792258315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004813372887656862,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19493725001811982,
      "num_tokens": 30622393.0,
      "step": 17650
    },
    {
      "entropy": 5.553914785385132,
      "epoch": 1.3736238085975492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004813262352876576,
      "loss": 5.1378,
      "mean_token_accuracy": 0.188840913772583,
      "num_tokens": 30632173.0,
      "step": 17655
    },
    {
      "entropy": 5.679634380340576,
      "epoch": 1.3740128379692667,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048131517867895684,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1797630175948143,
      "num_tokens": 30640313.0,
      "step": 17660
    },
    {
      "entropy": 5.596321964263916,
      "epoch": 1.3744018673409841,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048130411893975155,
      "loss": 5.0276,
      "mean_token_accuracy": 0.1961914226412773,
      "num_tokens": 30648651.0,
      "step": 17665
    },
    {
      "entropy": 5.562128305435181,
      "epoch": 1.3747908967127018,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004812930560702097,
      "loss": 5.1074,
      "mean_token_accuracy": 0.18621804118156432,
      "num_tokens": 30657782.0,
      "step": 17670
    },
    {
      "entropy": 5.580601835250855,
      "epoch": 1.3751799260844193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048128199007049907,
      "loss": 5.0988,
      "mean_token_accuracy": 0.1927732527256012,
      "num_tokens": 30666493.0,
      "step": 17675
    },
    {
      "entropy": 5.546082210540772,
      "epoch": 1.375568955456137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048127092094078756,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20225853472948074,
      "num_tokens": 30675054.0,
      "step": 17680
    },
    {
      "entropy": 5.6191143035888675,
      "epoch": 1.3759579848278545,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048125984868124317,
      "loss": 5.0225,
      "mean_token_accuracy": 0.20367906987667084,
      "num_tokens": 30683927.0,
      "step": 17685
    },
    {
      "entropy": 5.4911902904510494,
      "epoch": 1.3763470141995722,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004812487732920338,
      "loss": 4.9543,
      "mean_token_accuracy": 0.19853790253400802,
      "num_tokens": 30691670.0,
      "step": 17690
    },
    {
      "entropy": 5.505214071273803,
      "epoch": 1.3767360435712896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004812376947733277,
      "loss": 4.9826,
      "mean_token_accuracy": 0.20083227008581161,
      "num_tokens": 30700312.0,
      "step": 17695
    },
    {
      "entropy": 5.5463823795318605,
      "epoch": 1.3771250729430071,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048122661312529263,
      "loss": 5.029,
      "mean_token_accuracy": 0.20133189707994462,
      "num_tokens": 30708975.0,
      "step": 17700
    },
    {
      "entropy": 5.547616815567016,
      "epoch": 1.3775141023147248,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004812155283480971,
      "loss": 5.0742,
      "mean_token_accuracy": 0.1971016839146614,
      "num_tokens": 30718314.0,
      "step": 17705
    },
    {
      "entropy": 5.6133311748504635,
      "epoch": 1.3779031316864423,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004812044404419091,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19772496968507766,
      "num_tokens": 30726801.0,
      "step": 17710
    },
    {
      "entropy": 5.568340826034546,
      "epoch": 1.3782921610581598,
      "grad_norm": 1.078125,
      "learning_rate": 0.000481193349406897,
      "loss": 4.9877,
      "mean_token_accuracy": 0.19554893374443055,
      "num_tokens": 30734884.0,
      "step": 17715
    },
    {
      "entropy": 5.577547264099121,
      "epoch": 1.3786811904298775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000481182255243229,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19069030731916428,
      "num_tokens": 30744986.0,
      "step": 17720
    },
    {
      "entropy": 5.528880214691162,
      "epoch": 1.379070219801595,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048117115795107335,
      "loss": 5.0711,
      "mean_token_accuracy": 0.19110306054353715,
      "num_tokens": 30753476.0,
      "step": 17725
    },
    {
      "entropy": 5.645958375930786,
      "epoch": 1.3794592491733126,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004811600575305987,
      "loss": 5.1892,
      "mean_token_accuracy": 0.18294522613286973,
      "num_tokens": 30762642.0,
      "step": 17730
    },
    {
      "entropy": 5.685890245437622,
      "epoch": 1.3798482785450301,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004811489539819731,
      "loss": 5.1429,
      "mean_token_accuracy": 0.18279288709163666,
      "num_tokens": 30771530.0,
      "step": 17735
    },
    {
      "entropy": 5.619548273086548,
      "epoch": 1.3802373079167478,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048113784730536544,
      "loss": 5.141,
      "mean_token_accuracy": 0.19004807025194168,
      "num_tokens": 30780001.0,
      "step": 17740
    },
    {
      "entropy": 5.508500242233277,
      "epoch": 1.3806263372884653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048112673750094396,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19637584835290908,
      "num_tokens": 30788696.0,
      "step": 17745
    },
    {
      "entropy": 5.57595534324646,
      "epoch": 1.3810153666601828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004811156245688773,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20082703828811646,
      "num_tokens": 30797457.0,
      "step": 17750
    },
    {
      "entropy": 5.668876981735229,
      "epoch": 1.3814043960319005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004811045085093342,
      "loss": 5.0932,
      "mean_token_accuracy": 0.18886760920286177,
      "num_tokens": 30807048.0,
      "step": 17755
    },
    {
      "entropy": 5.595805501937866,
      "epoch": 1.381793425403618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004810933893224831,
      "loss": 5.1186,
      "mean_token_accuracy": 0.19311050027608873,
      "num_tokens": 30815315.0,
      "step": 17760
    },
    {
      "entropy": 5.554073905944824,
      "epoch": 1.3821824547753354,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048108226700849287,
      "loss": 5.0502,
      "mean_token_accuracy": 0.19583794325590134,
      "num_tokens": 30823221.0,
      "step": 17765
    },
    {
      "entropy": 5.633991241455078,
      "epoch": 1.382571484147053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004810711415675323,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19129905998706817,
      "num_tokens": 30832112.0,
      "step": 17770
    },
    {
      "entropy": 5.597090482711792,
      "epoch": 1.3829605135187706,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004810600129997702,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19559451192617416,
      "num_tokens": 30840597.0,
      "step": 17775
    },
    {
      "entropy": 5.554699420928955,
      "epoch": 1.3833495428904883,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048104888130537534,
      "loss": 4.916,
      "mean_token_accuracy": 0.20261192321777344,
      "num_tokens": 30848947.0,
      "step": 17780
    },
    {
      "entropy": 5.565868234634399,
      "epoch": 1.3837385722622058,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048103774648451664,
      "loss": 5.1326,
      "mean_token_accuracy": 0.18663860708475113,
      "num_tokens": 30856630.0,
      "step": 17785
    },
    {
      "entropy": 5.5948166847229,
      "epoch": 1.3841276016339235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048102660853736314,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1912068784236908,
      "num_tokens": 30866032.0,
      "step": 17790
    },
    {
      "entropy": 5.609870576858521,
      "epoch": 1.384516631005641,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004810154674640837,
      "loss": 5.0329,
      "mean_token_accuracy": 0.19983506500720977,
      "num_tokens": 30874104.0,
      "step": 17795
    },
    {
      "entropy": 5.574765205383301,
      "epoch": 1.3849056603773584,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048100432326484744,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20358753353357315,
      "num_tokens": 30882792.0,
      "step": 17800
    },
    {
      "entropy": 5.5719287395477295,
      "epoch": 1.385294689749076,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004809931759398235,
      "loss": 5.0276,
      "mean_token_accuracy": 0.19170805364847182,
      "num_tokens": 30891337.0,
      "step": 17805
    },
    {
      "entropy": 5.5796130180358885,
      "epoch": 1.3856837191207936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048098202548918094,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19743004441261292,
      "num_tokens": 30899589.0,
      "step": 17810
    },
    {
      "entropy": 5.621284866333008,
      "epoch": 1.386072748492511,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048097087191308914,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1891186609864235,
      "num_tokens": 30907893.0,
      "step": 17815
    },
    {
      "entropy": 5.66925573348999,
      "epoch": 1.3864617778642288,
      "grad_norm": 1.125,
      "learning_rate": 0.0004809597152117171,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1911892980337143,
      "num_tokens": 30916797.0,
      "step": 17820
    },
    {
      "entropy": 5.6509397506713865,
      "epoch": 1.3868508072359464,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004809485553852342,
      "loss": 5.0566,
      "mean_token_accuracy": 0.19857445359230042,
      "num_tokens": 30925451.0,
      "step": 17825
    },
    {
      "entropy": 5.573443841934204,
      "epoch": 1.387239836607664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004809373924338098,
      "loss": 5.0775,
      "mean_token_accuracy": 0.19146547764539718,
      "num_tokens": 30934595.0,
      "step": 17830
    },
    {
      "entropy": 5.598634672164917,
      "epoch": 1.3876288659793814,
      "grad_norm": 1.125,
      "learning_rate": 0.00048092622635761316,
      "loss": 5.1802,
      "mean_token_accuracy": 0.19262576401233672,
      "num_tokens": 30943896.0,
      "step": 17835
    },
    {
      "entropy": 5.633039760589599,
      "epoch": 1.388017895351099,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048091505715681395,
      "loss": 5.0233,
      "mean_token_accuracy": 0.19517601728439332,
      "num_tokens": 30952255.0,
      "step": 17840
    },
    {
      "entropy": 5.614893674850464,
      "epoch": 1.3884069247228166,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004809038848315814,
      "loss": 5.1182,
      "mean_token_accuracy": 0.1921895533800125,
      "num_tokens": 30960972.0,
      "step": 17845
    },
    {
      "entropy": 5.558756017684937,
      "epoch": 1.388795954094534,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004808927093820851,
      "loss": 5.1466,
      "mean_token_accuracy": 0.18826099038124083,
      "num_tokens": 30970298.0,
      "step": 17850
    },
    {
      "entropy": 5.643890762329102,
      "epoch": 1.3891849834662517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004808815308084947,
      "loss": 5.083,
      "mean_token_accuracy": 0.19906239658594133,
      "num_tokens": 30979355.0,
      "step": 17855
    },
    {
      "entropy": 5.5733216285705565,
      "epoch": 1.3895740128379692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004808703491109798,
      "loss": 5.0531,
      "mean_token_accuracy": 0.19875423312187196,
      "num_tokens": 30988085.0,
      "step": 17860
    },
    {
      "entropy": 5.57742486000061,
      "epoch": 1.3899630422096867,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004808591642897099,
      "loss": 5.046,
      "mean_token_accuracy": 0.1981653541326523,
      "num_tokens": 30996039.0,
      "step": 17865
    },
    {
      "entropy": 5.6551289558410645,
      "epoch": 1.3903520715814044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000480847976344855,
      "loss": 5.1823,
      "mean_token_accuracy": 0.18980445861816406,
      "num_tokens": 31004664.0,
      "step": 17870
    },
    {
      "entropy": 5.617430925369263,
      "epoch": 1.390741100953122,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048083678527658465,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1958012729883194,
      "num_tokens": 31012571.0,
      "step": 17875
    },
    {
      "entropy": 5.5749757289886475,
      "epoch": 1.3911301303248396,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004808255910850687,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19227755069732666,
      "num_tokens": 31020901.0,
      "step": 17880
    },
    {
      "entropy": 5.604028797149658,
      "epoch": 1.391519159696557,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004808143937704769,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19870197176933288,
      "num_tokens": 31029707.0,
      "step": 17885
    },
    {
      "entropy": 5.610916376113892,
      "epoch": 1.3919081890682747,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048080319333297937,
      "loss": 5.103,
      "mean_token_accuracy": 0.1951723501086235,
      "num_tokens": 31037833.0,
      "step": 17890
    },
    {
      "entropy": 5.591097164154053,
      "epoch": 1.3922972184399922,
      "grad_norm": 1.125,
      "learning_rate": 0.00048079198977274597,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20019507110118867,
      "num_tokens": 31046494.0,
      "step": 17895
    },
    {
      "entropy": 5.717887926101684,
      "epoch": 1.3926862478117097,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048078078308994666,
      "loss": 5.1671,
      "mean_token_accuracy": 0.1891368016600609,
      "num_tokens": 31054957.0,
      "step": 17900
    },
    {
      "entropy": 5.6580840110778805,
      "epoch": 1.3930752771834274,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004807695732847515,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19359701126813889,
      "num_tokens": 31063068.0,
      "step": 17905
    },
    {
      "entropy": 5.571873760223388,
      "epoch": 1.3934643065551449,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048075836035733053,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19341681152582169,
      "num_tokens": 31072081.0,
      "step": 17910
    },
    {
      "entropy": 5.589060926437378,
      "epoch": 1.3938533359268626,
      "grad_norm": 1.09375,
      "learning_rate": 0.000480747144307854,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20083021968603135,
      "num_tokens": 31080541.0,
      "step": 17915
    },
    {
      "entropy": 5.690802526473999,
      "epoch": 1.39424236529858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048073592513649203,
      "loss": 5.1622,
      "mean_token_accuracy": 0.18862955421209335,
      "num_tokens": 31089035.0,
      "step": 17920
    },
    {
      "entropy": 5.585626745223999,
      "epoch": 1.3946313946702977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004807247028434148,
      "loss": 5.1808,
      "mean_token_accuracy": 0.184421369433403,
      "num_tokens": 31096973.0,
      "step": 17925
    },
    {
      "entropy": 5.595784902572632,
      "epoch": 1.3950204240420152,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004807134774287927,
      "loss": 5.1117,
      "mean_token_accuracy": 0.18602322787046432,
      "num_tokens": 31104526.0,
      "step": 17930
    },
    {
      "entropy": 5.607011938095093,
      "epoch": 1.3954094534137327,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048070224889279603,
      "loss": 4.9558,
      "mean_token_accuracy": 0.20247727483510972,
      "num_tokens": 31112350.0,
      "step": 17935
    },
    {
      "entropy": 5.5938482761383055,
      "epoch": 1.3957984827854504,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048069101723559505,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1937098428606987,
      "num_tokens": 31120333.0,
      "step": 17940
    },
    {
      "entropy": 5.582429599761963,
      "epoch": 1.3961875121571679,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004806797824573603,
      "loss": 5.1226,
      "mean_token_accuracy": 0.1825736030936241,
      "num_tokens": 31129652.0,
      "step": 17945
    },
    {
      "entropy": 5.690742874145508,
      "epoch": 1.3965765415288853,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004806685445582624,
      "loss": 5.1114,
      "mean_token_accuracy": 0.19182732552289963,
      "num_tokens": 31137542.0,
      "step": 17950
    },
    {
      "entropy": 5.58651180267334,
      "epoch": 1.396965570900603,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048065730353847143,
      "loss": 5.0987,
      "mean_token_accuracy": 0.19038145691156388,
      "num_tokens": 31145958.0,
      "step": 17955
    },
    {
      "entropy": 5.570239686965943,
      "epoch": 1.3973546002723205,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048064605939815837,
      "loss": 5.0547,
      "mean_token_accuracy": 0.19258954972028733,
      "num_tokens": 31154267.0,
      "step": 17960
    },
    {
      "entropy": 5.664504432678223,
      "epoch": 1.3977436296440382,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004806348121374936,
      "loss": 5.1285,
      "mean_token_accuracy": 0.1940947562456131,
      "num_tokens": 31162742.0,
      "step": 17965
    },
    {
      "entropy": 5.66760892868042,
      "epoch": 1.3981326590157557,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004806235617566479,
      "loss": 5.1225,
      "mean_token_accuracy": 0.1966812089085579,
      "num_tokens": 31171275.0,
      "step": 17970
    },
    {
      "entropy": 5.529547166824341,
      "epoch": 1.3985216883874734,
      "grad_norm": 1.109375,
      "learning_rate": 0.000480612308255792,
      "loss": 4.8417,
      "mean_token_accuracy": 0.20422628968954087,
      "num_tokens": 31179318.0,
      "step": 17975
    },
    {
      "entropy": 5.585777330398559,
      "epoch": 1.3989107177591908,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048060105163509647,
      "loss": 5.0266,
      "mean_token_accuracy": 0.1991613268852234,
      "num_tokens": 31187129.0,
      "step": 17980
    },
    {
      "entropy": 5.624492931365967,
      "epoch": 1.3992997471309083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004805897918947323,
      "loss": 5.2202,
      "mean_token_accuracy": 0.18666080832481385,
      "num_tokens": 31196381.0,
      "step": 17985
    },
    {
      "entropy": 5.656292867660523,
      "epoch": 1.399688776502626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004805785290348702,
      "loss": 5.1431,
      "mean_token_accuracy": 0.18882026076316832,
      "num_tokens": 31205659.0,
      "step": 17990
    },
    {
      "entropy": 5.624086475372314,
      "epoch": 1.4000778058743435,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048056726305568124,
      "loss": 5.1103,
      "mean_token_accuracy": 0.19290167689323426,
      "num_tokens": 31214996.0,
      "step": 17995
    },
    {
      "entropy": 5.658321952819824,
      "epoch": 1.400466835246061,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048055599395733617,
      "loss": 5.0725,
      "mean_token_accuracy": 0.198309825360775,
      "num_tokens": 31222964.0,
      "step": 18000
    },
    {
      "epoch": 1.400466835246061,
      "eval_entropy": 5.370768020104857,
      "eval_loss": 5.1569695472717285,
      "eval_mean_token_accuracy": 0.20003243950522784,
      "eval_num_tokens": 31222964.0,
      "eval_runtime": 21.9651,
      "eval_samples_per_second": 1892.002,
      "eval_steps_per_second": 236.512,
      "step": 18000
    },
    {
      "entropy": 5.623855400085449,
      "epoch": 1.4008558646177787,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048054472174000615,
      "loss": 5.1909,
      "mean_token_accuracy": 0.18868928104639054,
      "num_tokens": 31231653.0,
      "step": 18005
    },
    {
      "entropy": 5.57370958328247,
      "epoch": 1.4012448939894961,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048053344640386207,
      "loss": 5.0499,
      "mean_token_accuracy": 0.1935676395893097,
      "num_tokens": 31240613.0,
      "step": 18010
    },
    {
      "entropy": 5.526791000366211,
      "epoch": 1.4016339233612138,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048052216794907505,
      "loss": 5.0034,
      "mean_token_accuracy": 0.19510119110345842,
      "num_tokens": 31249724.0,
      "step": 18015
    },
    {
      "entropy": 5.6243163585662845,
      "epoch": 1.4020229527329313,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048051088637581624,
      "loss": 5.2031,
      "mean_token_accuracy": 0.18628655076026918,
      "num_tokens": 31258717.0,
      "step": 18020
    },
    {
      "entropy": 5.596420431137085,
      "epoch": 1.402411982104649,
      "grad_norm": 1.0,
      "learning_rate": 0.000480499601684257,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20009602904319762,
      "num_tokens": 31267170.0,
      "step": 18025
    },
    {
      "entropy": 5.568233251571655,
      "epoch": 1.4028010114763665,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004804883138745682,
      "loss": 5.1207,
      "mean_token_accuracy": 0.1963480994105339,
      "num_tokens": 31277787.0,
      "step": 18030
    },
    {
      "entropy": 5.602793884277344,
      "epoch": 1.403190040848084,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004804770229469214,
      "loss": 5.1019,
      "mean_token_accuracy": 0.18460707664489745,
      "num_tokens": 31286185.0,
      "step": 18035
    },
    {
      "entropy": 5.634204626083374,
      "epoch": 1.4035790702198017,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004804657289014878,
      "loss": 5.0266,
      "mean_token_accuracy": 0.19173173755407333,
      "num_tokens": 31295361.0,
      "step": 18040
    },
    {
      "entropy": 5.598345804214477,
      "epoch": 1.4039680995915191,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048045443173843884,
      "loss": 5.1296,
      "mean_token_accuracy": 0.1976161226630211,
      "num_tokens": 31304036.0,
      "step": 18045
    },
    {
      "entropy": 5.516292762756348,
      "epoch": 1.4043571289632366,
      "grad_norm": 1.125,
      "learning_rate": 0.00048044313145794587,
      "loss": 5.0287,
      "mean_token_accuracy": 0.19602672457695008,
      "num_tokens": 31311919.0,
      "step": 18050
    },
    {
      "entropy": 5.675953435897827,
      "epoch": 1.4047461583349543,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048043182806018034,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1873700276017189,
      "num_tokens": 31320805.0,
      "step": 18055
    },
    {
      "entropy": 5.662509918212891,
      "epoch": 1.4051351877066718,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048042052154531384,
      "loss": 5.1377,
      "mean_token_accuracy": 0.19272580742835999,
      "num_tokens": 31329688.0,
      "step": 18060
    },
    {
      "entropy": 5.54470853805542,
      "epoch": 1.4055242170783895,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004804092119135179,
      "loss": 4.9981,
      "mean_token_accuracy": 0.1981380432844162,
      "num_tokens": 31338301.0,
      "step": 18065
    },
    {
      "entropy": 5.669227552413941,
      "epoch": 1.405913246450107,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004803978991649641,
      "loss": 5.1807,
      "mean_token_accuracy": 0.18489658534526826,
      "num_tokens": 31347371.0,
      "step": 18070
    },
    {
      "entropy": 5.678305959701538,
      "epoch": 1.4063022758218247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048038658329982404,
      "loss": 5.1228,
      "mean_token_accuracy": 0.19393701404333114,
      "num_tokens": 31355740.0,
      "step": 18075
    },
    {
      "entropy": 5.548440265655517,
      "epoch": 1.4066913051935421,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004803752643182695,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20297455042600632,
      "num_tokens": 31364718.0,
      "step": 18080
    },
    {
      "entropy": 5.680179309844971,
      "epoch": 1.4070803345652596,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004803639422204723,
      "loss": 5.113,
      "mean_token_accuracy": 0.19262794852256776,
      "num_tokens": 31372845.0,
      "step": 18085
    },
    {
      "entropy": 5.641577482223511,
      "epoch": 1.4074693639369773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004803526170066041,
      "loss": 5.1097,
      "mean_token_accuracy": 0.1924801766872406,
      "num_tokens": 31380614.0,
      "step": 18090
    },
    {
      "entropy": 5.659421062469482,
      "epoch": 1.4078583933086948,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048034128867683674,
      "loss": 5.1703,
      "mean_token_accuracy": 0.18953775614500046,
      "num_tokens": 31389609.0,
      "step": 18095
    },
    {
      "entropy": 5.631828451156617,
      "epoch": 1.4082474226804123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004803299572313422,
      "loss": 5.131,
      "mean_token_accuracy": 0.19067511782050134,
      "num_tokens": 31399046.0,
      "step": 18100
    },
    {
      "entropy": 5.578372621536255,
      "epoch": 1.40863645205213,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004803186226702924,
      "loss": 5.0552,
      "mean_token_accuracy": 0.19935795217752456,
      "num_tokens": 31407832.0,
      "step": 18105
    },
    {
      "entropy": 5.608093547821045,
      "epoch": 1.4090254814238474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004803072849938592,
      "loss": 5.1262,
      "mean_token_accuracy": 0.18821159154176711,
      "num_tokens": 31417097.0,
      "step": 18110
    },
    {
      "entropy": 5.642530345916748,
      "epoch": 1.4094145107955651,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004802959442022149,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1884885013103485,
      "num_tokens": 31425448.0,
      "step": 18115
    },
    {
      "entropy": 5.551266098022461,
      "epoch": 1.4098035401672826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048028460029553126,
      "loss": 4.9788,
      "mean_token_accuracy": 0.2092061772942543,
      "num_tokens": 31434116.0,
      "step": 18120
    },
    {
      "entropy": 5.558471965789795,
      "epoch": 1.4101925695390003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048027325327398065,
      "loss": 5.0544,
      "mean_token_accuracy": 0.191595421731472,
      "num_tokens": 31443456.0,
      "step": 18125
    },
    {
      "entropy": 5.599004888534546,
      "epoch": 1.4105815989107178,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004802619031377351,
      "loss": 5.0545,
      "mean_token_accuracy": 0.19618366211652755,
      "num_tokens": 31452372.0,
      "step": 18130
    },
    {
      "entropy": 5.6977051258087155,
      "epoch": 1.4109706282824352,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048025054988696684,
      "loss": 5.0981,
      "mean_token_accuracy": 0.2002044662833214,
      "num_tokens": 31460446.0,
      "step": 18135
    },
    {
      "entropy": 5.594892835617065,
      "epoch": 1.411359657654153,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048023919352184827,
      "loss": 5.0552,
      "mean_token_accuracy": 0.18927150964736938,
      "num_tokens": 31468814.0,
      "step": 18140
    },
    {
      "entropy": 5.53581657409668,
      "epoch": 1.4117486870258704,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048022783404255156,
      "loss": 5.0106,
      "mean_token_accuracy": 0.1957953691482544,
      "num_tokens": 31477617.0,
      "step": 18145
    },
    {
      "entropy": 5.586964797973633,
      "epoch": 1.412137716397588,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004802164714492491,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20209989100694656,
      "num_tokens": 31485226.0,
      "step": 18150
    },
    {
      "entropy": 5.589772415161133,
      "epoch": 1.4125267457693056,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048020510574211335,
      "loss": 5.0428,
      "mean_token_accuracy": 0.19675385802984238,
      "num_tokens": 31493425.0,
      "step": 18155
    },
    {
      "entropy": 5.56939377784729,
      "epoch": 1.412915775141023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048019373692131674,
      "loss": 5.0064,
      "mean_token_accuracy": 0.20103743076324462,
      "num_tokens": 31502499.0,
      "step": 18160
    },
    {
      "entropy": 5.611530017852783,
      "epoch": 1.4133048045127408,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048018236498703176,
      "loss": 5.0676,
      "mean_token_accuracy": 0.19808191955089569,
      "num_tokens": 31510376.0,
      "step": 18165
    },
    {
      "entropy": 5.677847766876221,
      "epoch": 1.4136938338844582,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048017098993943097,
      "loss": 5.1984,
      "mean_token_accuracy": 0.1899036943912506,
      "num_tokens": 31519101.0,
      "step": 18170
    },
    {
      "entropy": 5.59563102722168,
      "epoch": 1.414082863256176,
      "grad_norm": 1.15625,
      "learning_rate": 0.000480159611778687,
      "loss": 5.0276,
      "mean_token_accuracy": 0.2008332848548889,
      "num_tokens": 31527221.0,
      "step": 18175
    },
    {
      "entropy": 5.580407905578613,
      "epoch": 1.4144718926278934,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048014823050497243,
      "loss": 5.073,
      "mean_token_accuracy": 0.19431888163089753,
      "num_tokens": 31535711.0,
      "step": 18180
    },
    {
      "entropy": 5.660232162475586,
      "epoch": 1.4148609219996109,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048013684611846,
      "loss": 5.2083,
      "mean_token_accuracy": 0.18721220940351485,
      "num_tokens": 31545363.0,
      "step": 18185
    },
    {
      "entropy": 5.6775195598602295,
      "epoch": 1.4152499513713286,
      "grad_norm": 1.125,
      "learning_rate": 0.00048012545861932245,
      "loss": 5.0639,
      "mean_token_accuracy": 0.1999210983514786,
      "num_tokens": 31553760.0,
      "step": 18190
    },
    {
      "entropy": 5.592566585540771,
      "epoch": 1.415638980743046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004801140680077325,
      "loss": 5.014,
      "mean_token_accuracy": 0.19732030481100082,
      "num_tokens": 31561625.0,
      "step": 18195
    },
    {
      "entropy": 5.554940605163575,
      "epoch": 1.4160280101147635,
      "grad_norm": 1.125,
      "learning_rate": 0.0004801026742838631,
      "loss": 5.1044,
      "mean_token_accuracy": 0.19561590254306793,
      "num_tokens": 31569345.0,
      "step": 18200
    },
    {
      "entropy": 5.623496723175049,
      "epoch": 1.4164170394864812,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004800912774478871,
      "loss": 5.0458,
      "mean_token_accuracy": 0.19155776500701904,
      "num_tokens": 31577990.0,
      "step": 18205
    },
    {
      "entropy": 5.589687252044678,
      "epoch": 1.416806068858199,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004800798774999773,
      "loss": 5.011,
      "mean_token_accuracy": 0.19557437300682068,
      "num_tokens": 31586556.0,
      "step": 18210
    },
    {
      "entropy": 5.641947507858276,
      "epoch": 1.4171950982299164,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048006847444030686,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19611188173294067,
      "num_tokens": 31595697.0,
      "step": 18215
    },
    {
      "entropy": 5.5587756633758545,
      "epoch": 1.4175841276016339,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004800570682690487,
      "loss": 5.0252,
      "mean_token_accuracy": 0.1990537703037262,
      "num_tokens": 31604760.0,
      "step": 18220
    },
    {
      "entropy": 5.636505222320556,
      "epoch": 1.4179731569733516,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004800456589863759,
      "loss": 5.1873,
      "mean_token_accuracy": 0.18376110643148422,
      "num_tokens": 31614115.0,
      "step": 18225
    },
    {
      "entropy": 5.614505672454834,
      "epoch": 1.418362186345069,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004800342465924616,
      "loss": 5.1489,
      "mean_token_accuracy": 0.189004522562027,
      "num_tokens": 31622492.0,
      "step": 18230
    },
    {
      "entropy": 5.640391540527344,
      "epoch": 1.4187512157167865,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000480022831087479,
      "loss": 5.1495,
      "mean_token_accuracy": 0.18951364308595658,
      "num_tokens": 31632269.0,
      "step": 18235
    },
    {
      "entropy": 5.630290460586548,
      "epoch": 1.4191402450885042,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004800114124716012,
      "loss": 5.1358,
      "mean_token_accuracy": 0.19162488579750062,
      "num_tokens": 31640374.0,
      "step": 18240
    },
    {
      "entropy": 5.600970697402954,
      "epoch": 1.4195292744602217,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004799999907450015,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2009406492114067,
      "num_tokens": 31648406.0,
      "step": 18245
    },
    {
      "entropy": 5.584198474884033,
      "epoch": 1.4199183038319392,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004799885659078533,
      "loss": 5.0348,
      "mean_token_accuracy": 0.1939269036054611,
      "num_tokens": 31656478.0,
      "step": 18250
    },
    {
      "entropy": 5.599806880950927,
      "epoch": 1.4203073332036569,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004799771379603299,
      "loss": 5.0358,
      "mean_token_accuracy": 0.1918930158019066,
      "num_tokens": 31665106.0,
      "step": 18255
    },
    {
      "entropy": 5.66982479095459,
      "epoch": 1.4206963625753746,
      "grad_norm": 1.125,
      "learning_rate": 0.0004799657069026046,
      "loss": 5.1349,
      "mean_token_accuracy": 0.19132102727890016,
      "num_tokens": 31673557.0,
      "step": 18260
    },
    {
      "entropy": 5.553587532043457,
      "epoch": 1.421085391947092,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047995427273485097,
      "loss": 4.9583,
      "mean_token_accuracy": 0.19654569923877716,
      "num_tokens": 31682123.0,
      "step": 18265
    },
    {
      "entropy": 5.653144025802613,
      "epoch": 1.4214744213188095,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047994283545724247,
      "loss": 5.2236,
      "mean_token_accuracy": 0.1842864751815796,
      "num_tokens": 31690387.0,
      "step": 18270
    },
    {
      "entropy": 5.5994140148162845,
      "epoch": 1.4218634506905272,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047993139506995273,
      "loss": 5.0322,
      "mean_token_accuracy": 0.20648835152387618,
      "num_tokens": 31699814.0,
      "step": 18275
    },
    {
      "entropy": 5.554902124404907,
      "epoch": 1.4222524800622447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004799199515731551,
      "loss": 5.015,
      "mean_token_accuracy": 0.20689452439546585,
      "num_tokens": 31708437.0,
      "step": 18280
    },
    {
      "entropy": 5.566213798522949,
      "epoch": 1.4226415094339622,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047990850496702346,
      "loss": 4.985,
      "mean_token_accuracy": 0.20756625086069108,
      "num_tokens": 31716935.0,
      "step": 18285
    },
    {
      "entropy": 5.696743631362915,
      "epoch": 1.4230305388056799,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004798970552517314,
      "loss": 5.2102,
      "mean_token_accuracy": 0.18929355591535568,
      "num_tokens": 31725997.0,
      "step": 18290
    },
    {
      "entropy": 5.632201862335205,
      "epoch": 1.4234195681773973,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047988560242745264,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19838460236787797,
      "num_tokens": 31733651.0,
      "step": 18295
    },
    {
      "entropy": 5.563215446472168,
      "epoch": 1.423808597549115,
      "grad_norm": 1.15625,
      "learning_rate": 0.000479874146494361,
      "loss": 5.033,
      "mean_token_accuracy": 0.195961694419384,
      "num_tokens": 31742465.0,
      "step": 18300
    },
    {
      "entropy": 5.5926672458648685,
      "epoch": 1.4241976269208325,
      "grad_norm": 1.125,
      "learning_rate": 0.0004798626874526302,
      "loss": 5.0954,
      "mean_token_accuracy": 0.19464299827814102,
      "num_tokens": 31751547.0,
      "step": 18305
    },
    {
      "entropy": 5.66100926399231,
      "epoch": 1.4245866562925502,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047985122530243426,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1839061975479126,
      "num_tokens": 31761027.0,
      "step": 18310
    },
    {
      "entropy": 5.667470788955688,
      "epoch": 1.4249756856642677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000479839760043947,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1856790989637375,
      "num_tokens": 31770166.0,
      "step": 18315
    },
    {
      "entropy": 5.641820669174194,
      "epoch": 1.4253647150359852,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047982829167734245,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2043808728456497,
      "num_tokens": 31778606.0,
      "step": 18320
    },
    {
      "entropy": 5.640340566635132,
      "epoch": 1.4257537444077029,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047981682020279456,
      "loss": 5.2247,
      "mean_token_accuracy": 0.18577693849802018,
      "num_tokens": 31787637.0,
      "step": 18325
    },
    {
      "entropy": 5.558642911911011,
      "epoch": 1.4261427737794203,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004798053456204775,
      "loss": 5.0191,
      "mean_token_accuracy": 0.20263041853904723,
      "num_tokens": 31796401.0,
      "step": 18330
    },
    {
      "entropy": 5.663474082946777,
      "epoch": 1.4265318031511378,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004797938679305651,
      "loss": 5.1572,
      "mean_token_accuracy": 0.18735083490610122,
      "num_tokens": 31805917.0,
      "step": 18335
    },
    {
      "entropy": 5.701756811141967,
      "epoch": 1.4269208325228555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047978238713323193,
      "loss": 5.1564,
      "mean_token_accuracy": 0.18942782878875733,
      "num_tokens": 31814890.0,
      "step": 18340
    },
    {
      "entropy": 5.645923185348511,
      "epoch": 1.427309861894573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047977090322865183,
      "loss": 5.0065,
      "mean_token_accuracy": 0.1999688595533371,
      "num_tokens": 31822469.0,
      "step": 18345
    },
    {
      "entropy": 5.5979905128479,
      "epoch": 1.4276988912662907,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004797594162169993,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19087280333042145,
      "num_tokens": 31831030.0,
      "step": 18350
    },
    {
      "entropy": 5.587630033493042,
      "epoch": 1.4280879206380082,
      "grad_norm": 1.125,
      "learning_rate": 0.0004797479260984485,
      "loss": 5.0534,
      "mean_token_accuracy": 0.19322826117277145,
      "num_tokens": 31840137.0,
      "step": 18355
    },
    {
      "entropy": 5.52805461883545,
      "epoch": 1.4284769500097259,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047973643287317386,
      "loss": 4.9605,
      "mean_token_accuracy": 0.20177684873342513,
      "num_tokens": 31849258.0,
      "step": 18360
    },
    {
      "entropy": 5.564167404174805,
      "epoch": 1.4288659793814433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047972493654134963,
      "loss": 4.9819,
      "mean_token_accuracy": 0.19464049935340882,
      "num_tokens": 31857506.0,
      "step": 18365
    },
    {
      "entropy": 5.613288116455078,
      "epoch": 1.4292550087531608,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047971343710315043,
      "loss": 5.0811,
      "mean_token_accuracy": 0.19314426630735398,
      "num_tokens": 31866685.0,
      "step": 18370
    },
    {
      "entropy": 5.681535148620606,
      "epoch": 1.4296440381248785,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004797019345587506,
      "loss": 5.1792,
      "mean_token_accuracy": 0.18092888742685317,
      "num_tokens": 31874404.0,
      "step": 18375
    },
    {
      "entropy": 5.651321220397949,
      "epoch": 1.430033067496596,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004796904289083248,
      "loss": 5.1464,
      "mean_token_accuracy": 0.1945211783051491,
      "num_tokens": 31883075.0,
      "step": 18380
    },
    {
      "entropy": 5.574392461776734,
      "epoch": 1.4304220968683135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004796789201520474,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19601092636585235,
      "num_tokens": 31891539.0,
      "step": 18385
    },
    {
      "entropy": 5.655614280700684,
      "epoch": 1.4308111262400312,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047966740829009333,
      "loss": 5.1153,
      "mean_token_accuracy": 0.1918170765042305,
      "num_tokens": 31900575.0,
      "step": 18390
    },
    {
      "entropy": 5.6333899974823,
      "epoch": 1.4312001556117486,
      "grad_norm": 1.0625,
      "learning_rate": 0.000479655893322637,
      "loss": 5.0782,
      "mean_token_accuracy": 0.1922677606344223,
      "num_tokens": 31909256.0,
      "step": 18395
    },
    {
      "entropy": 5.550563144683838,
      "epoch": 1.4315891849834663,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004796443752498532,
      "loss": 5.03,
      "mean_token_accuracy": 0.1974807173013687,
      "num_tokens": 31917098.0,
      "step": 18400
    },
    {
      "entropy": 5.702553749084473,
      "epoch": 1.4319782143551838,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004796328540719169,
      "loss": 5.1638,
      "mean_token_accuracy": 0.19215747267007827,
      "num_tokens": 31925331.0,
      "step": 18405
    },
    {
      "entropy": 5.627672958374023,
      "epoch": 1.4323672437269015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004796213297890026,
      "loss": 5.0379,
      "mean_token_accuracy": 0.1983242392539978,
      "num_tokens": 31933769.0,
      "step": 18410
    },
    {
      "entropy": 5.626649808883667,
      "epoch": 1.432756273098619,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004796098024012853,
      "loss": 5.1277,
      "mean_token_accuracy": 0.18745720535516738,
      "num_tokens": 31943089.0,
      "step": 18415
    },
    {
      "entropy": 5.635249757766724,
      "epoch": 1.4331453024703364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047959827190894,
      "loss": 5.0803,
      "mean_token_accuracy": 0.19523430466651917,
      "num_tokens": 31951452.0,
      "step": 18420
    },
    {
      "entropy": 5.655068635940552,
      "epoch": 1.4335343318420541,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047958673831214157,
      "loss": 5.1144,
      "mean_token_accuracy": 0.19457911252975463,
      "num_tokens": 31959625.0,
      "step": 18425
    },
    {
      "entropy": 5.554893541336059,
      "epoch": 1.4339233612137716,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047957520161106496,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19713234901428223,
      "num_tokens": 31967906.0,
      "step": 18430
    },
    {
      "entropy": 5.548131704330444,
      "epoch": 1.434312390585489,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004795636618058853,
      "loss": 5.0369,
      "mean_token_accuracy": 0.19901105612516404,
      "num_tokens": 31976183.0,
      "step": 18435
    },
    {
      "entropy": 5.5955760955810545,
      "epoch": 1.4347014199572068,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004795521188967777,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2000845581293106,
      "num_tokens": 31984513.0,
      "step": 18440
    },
    {
      "entropy": 5.728031826019287,
      "epoch": 1.4350904493289243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004795405728839172,
      "loss": 5.2169,
      "mean_token_accuracy": 0.1833504021167755,
      "num_tokens": 31992994.0,
      "step": 18445
    },
    {
      "entropy": 5.668539190292359,
      "epoch": 1.435479478700642,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004795290237674792,
      "loss": 5.078,
      "mean_token_accuracy": 0.19417960494756697,
      "num_tokens": 32002043.0,
      "step": 18450
    },
    {
      "entropy": 5.603974390029907,
      "epoch": 1.4358685080723594,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047951747154763866,
      "loss": 5.0173,
      "mean_token_accuracy": 0.19747249335050582,
      "num_tokens": 32009901.0,
      "step": 18455
    },
    {
      "entropy": 5.577097415924072,
      "epoch": 1.4362575374440771,
      "grad_norm": 1.125,
      "learning_rate": 0.00047950591622457106,
      "loss": 5.047,
      "mean_token_accuracy": 0.19376734644174576,
      "num_tokens": 32017947.0,
      "step": 18460
    },
    {
      "entropy": 5.6251757621765135,
      "epoch": 1.4366465668157946,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047949435779845174,
      "loss": 5.0816,
      "mean_token_accuracy": 0.18930541425943376,
      "num_tokens": 32026537.0,
      "step": 18465
    },
    {
      "entropy": 5.619577360153198,
      "epoch": 1.437035596187512,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047948279626945596,
      "loss": 5.0307,
      "mean_token_accuracy": 0.19658903479576112,
      "num_tokens": 32034874.0,
      "step": 18470
    },
    {
      "entropy": 5.609281396865844,
      "epoch": 1.4374246255592298,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047947123163775924,
      "loss": 5.1354,
      "mean_token_accuracy": 0.18824116438627242,
      "num_tokens": 32043829.0,
      "step": 18475
    },
    {
      "entropy": 5.6359399318695065,
      "epoch": 1.4378136549309473,
      "grad_norm": 1.046875,
      "learning_rate": 0.000479459663903537,
      "loss": 5.1608,
      "mean_token_accuracy": 0.185589037835598,
      "num_tokens": 32052883.0,
      "step": 18480
    },
    {
      "entropy": 5.6727738857269285,
      "epoch": 1.4382026843026647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004794480930669649,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19992732405662536,
      "num_tokens": 32061421.0,
      "step": 18485
    },
    {
      "entropy": 5.626596164703369,
      "epoch": 1.4385917136743824,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004794365191282183,
      "loss": 5.082,
      "mean_token_accuracy": 0.20136156529188157,
      "num_tokens": 32069957.0,
      "step": 18490
    },
    {
      "entropy": 5.5335307121276855,
      "epoch": 1.4389807430461,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047942494208747297,
      "loss": 4.9869,
      "mean_token_accuracy": 0.19786201268434525,
      "num_tokens": 32079375.0,
      "step": 18495
    },
    {
      "entropy": 5.539895868301391,
      "epoch": 1.4393697724178176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004794133619449045,
      "loss": 5.0194,
      "mean_token_accuracy": 0.19809168726205825,
      "num_tokens": 32088331.0,
      "step": 18500
    },
    {
      "entropy": 5.588575649261474,
      "epoch": 1.439758801789535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004794017787006886,
      "loss": 4.978,
      "mean_token_accuracy": 0.1992468237876892,
      "num_tokens": 32096897.0,
      "step": 18505
    },
    {
      "entropy": 5.554288291931153,
      "epoch": 1.4401478311612528,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000479390192355001,
      "loss": 5.015,
      "mean_token_accuracy": 0.2040394812822342,
      "num_tokens": 32105942.0,
      "step": 18510
    },
    {
      "entropy": 5.665124750137329,
      "epoch": 1.4405368605329703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004793786029080176,
      "loss": 5.213,
      "mean_token_accuracy": 0.1890985459089279,
      "num_tokens": 32114540.0,
      "step": 18515
    },
    {
      "entropy": 5.566246175765992,
      "epoch": 1.4409258899046877,
      "grad_norm": 1.125,
      "learning_rate": 0.0004793670103599143,
      "loss": 5.0367,
      "mean_token_accuracy": 0.1951267197728157,
      "num_tokens": 32123214.0,
      "step": 18520
    },
    {
      "entropy": 5.5863861560821535,
      "epoch": 1.4413149192764054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047935541471086677,
      "loss": 5.0787,
      "mean_token_accuracy": 0.1991425171494484,
      "num_tokens": 32132369.0,
      "step": 18525
    },
    {
      "entropy": 5.5568053245544435,
      "epoch": 1.441703948648123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004793438159610511,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1981082171201706,
      "num_tokens": 32141037.0,
      "step": 18530
    },
    {
      "entropy": 5.607669544219971,
      "epoch": 1.4420929780198404,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047933221411064334,
      "loss": 5.0472,
      "mean_token_accuracy": 0.1980926901102066,
      "num_tokens": 32149515.0,
      "step": 18535
    },
    {
      "entropy": 5.593973922729492,
      "epoch": 1.442482007391558,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004793206091598194,
      "loss": 5.117,
      "mean_token_accuracy": 0.20115037709474565,
      "num_tokens": 32158099.0,
      "step": 18540
    },
    {
      "entropy": 5.608628988265991,
      "epoch": 1.4428710367632755,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004793090011087554,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19916650652885437,
      "num_tokens": 32167059.0,
      "step": 18545
    },
    {
      "entropy": 5.681153249740601,
      "epoch": 1.4432600661349932,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047929738995762755,
      "loss": 5.2386,
      "mean_token_accuracy": 0.18743796795606613,
      "num_tokens": 32175988.0,
      "step": 18550
    },
    {
      "entropy": 5.559827852249145,
      "epoch": 1.4436490955067107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000479285775706612,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19665402173995972,
      "num_tokens": 32185362.0,
      "step": 18555
    },
    {
      "entropy": 5.582279920578003,
      "epoch": 1.4440381248784284,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047927415835588496,
      "loss": 5.0566,
      "mean_token_accuracy": 0.1985292226076126,
      "num_tokens": 32194599.0,
      "step": 18560
    },
    {
      "entropy": 5.635741472244263,
      "epoch": 1.444427154250146,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047926253790562265,
      "loss": 5.0603,
      "mean_token_accuracy": 0.19462480992078782,
      "num_tokens": 32203106.0,
      "step": 18565
    },
    {
      "entropy": 5.59201865196228,
      "epoch": 1.4448161836218634,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047925091435600147,
      "loss": 5.1161,
      "mean_token_accuracy": 0.19246195405721664,
      "num_tokens": 32211652.0,
      "step": 18570
    },
    {
      "entropy": 5.675584220886231,
      "epoch": 1.445205212993581,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047923928770719776,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19302731156349182,
      "num_tokens": 32220435.0,
      "step": 18575
    },
    {
      "entropy": 5.70727653503418,
      "epoch": 1.4455942423652985,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047922765795938797,
      "loss": 5.0487,
      "mean_token_accuracy": 0.19232486188411713,
      "num_tokens": 32228368.0,
      "step": 18580
    },
    {
      "entropy": 5.579379510879517,
      "epoch": 1.445983271737016,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004792160251127485,
      "loss": 5.1126,
      "mean_token_accuracy": 0.18917977958917617,
      "num_tokens": 32237367.0,
      "step": 18585
    },
    {
      "entropy": 5.612540245056152,
      "epoch": 1.4463723011087337,
      "grad_norm": 1.125,
      "learning_rate": 0.00047920438916745586,
      "loss": 5.1134,
      "mean_token_accuracy": 0.1923543930053711,
      "num_tokens": 32245503.0,
      "step": 18590
    },
    {
      "entropy": 5.686342477798462,
      "epoch": 1.4467613304804512,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004791927501236866,
      "loss": 5.1025,
      "mean_token_accuracy": 0.1902190014719963,
      "num_tokens": 32254565.0,
      "step": 18595
    },
    {
      "entropy": 5.645481586456299,
      "epoch": 1.4471503598521689,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047918110798161754,
      "loss": 5.1349,
      "mean_token_accuracy": 0.18823806345462799,
      "num_tokens": 32263701.0,
      "step": 18600
    },
    {
      "entropy": 5.55140528678894,
      "epoch": 1.4475393892238864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047916946274142503,
      "loss": 5.0817,
      "mean_token_accuracy": 0.193499393761158,
      "num_tokens": 32272205.0,
      "step": 18605
    },
    {
      "entropy": 5.646130037307739,
      "epoch": 1.447928418595604,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047915781440328593,
      "loss": 5.124,
      "mean_token_accuracy": 0.1942844957113266,
      "num_tokens": 32279918.0,
      "step": 18610
    },
    {
      "entropy": 5.65764627456665,
      "epoch": 1.4483174479673215,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004791461629673769,
      "loss": 5.12,
      "mean_token_accuracy": 0.19900160282850266,
      "num_tokens": 32288600.0,
      "step": 18615
    },
    {
      "entropy": 5.6804746150970455,
      "epoch": 1.448706477339039,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004791345084338748,
      "loss": 5.2077,
      "mean_token_accuracy": 0.19376307725906372,
      "num_tokens": 32298334.0,
      "step": 18620
    },
    {
      "entropy": 5.613220882415772,
      "epoch": 1.4490955067107567,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004791228508029565,
      "loss": 5.0054,
      "mean_token_accuracy": 0.1985485166311264,
      "num_tokens": 32306785.0,
      "step": 18625
    },
    {
      "entropy": 5.606794786453247,
      "epoch": 1.4494845360824742,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047911119007479873,
      "loss": 5.1676,
      "mean_token_accuracy": 0.19212428629398345,
      "num_tokens": 32315419.0,
      "step": 18630
    },
    {
      "entropy": 5.622725868225098,
      "epoch": 1.4498735654541917,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047909952624957866,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20365918278694153,
      "num_tokens": 32323655.0,
      "step": 18635
    },
    {
      "entropy": 5.580766868591309,
      "epoch": 1.4502625948259094,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000479087859327473,
      "loss": 5.0114,
      "mean_token_accuracy": 0.19585212469100952,
      "num_tokens": 32331465.0,
      "step": 18640
    },
    {
      "entropy": 5.65385627746582,
      "epoch": 1.450651624197627,
      "grad_norm": 1.09375,
      "learning_rate": 0.000479076189308659,
      "loss": 5.1978,
      "mean_token_accuracy": 0.18625448942184447,
      "num_tokens": 32341211.0,
      "step": 18645
    },
    {
      "entropy": 5.672112846374512,
      "epoch": 1.4510406535693445,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047906451619331364,
      "loss": 5.2315,
      "mean_token_accuracy": 0.1861101359128952,
      "num_tokens": 32349344.0,
      "step": 18650
    },
    {
      "entropy": 5.6738330841064455,
      "epoch": 1.451429682941062,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047905283998161404,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1998172104358673,
      "num_tokens": 32357797.0,
      "step": 18655
    },
    {
      "entropy": 5.543853282928467,
      "epoch": 1.4518187123127797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047904116067373743,
      "loss": 5.0132,
      "mean_token_accuracy": 0.19805641919374467,
      "num_tokens": 32366131.0,
      "step": 18660
    },
    {
      "entropy": 5.583631896972657,
      "epoch": 1.4522077416844972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004790294782698609,
      "loss": 5.1555,
      "mean_token_accuracy": 0.18505419045686722,
      "num_tokens": 32374738.0,
      "step": 18665
    },
    {
      "entropy": 5.669122362136841,
      "epoch": 1.4525967710562147,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004790177927701618,
      "loss": 5.1049,
      "mean_token_accuracy": 0.1930384784936905,
      "num_tokens": 32383638.0,
      "step": 18670
    },
    {
      "entropy": 5.665137434005738,
      "epoch": 1.4529858004279324,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004790061041748174,
      "loss": 5.1473,
      "mean_token_accuracy": 0.18912722468376159,
      "num_tokens": 32393269.0,
      "step": 18675
    },
    {
      "entropy": 5.655826044082642,
      "epoch": 1.4533748297996498,
      "grad_norm": 1.125,
      "learning_rate": 0.0004789944124840051,
      "loss": 5.1037,
      "mean_token_accuracy": 0.19152767211198807,
      "num_tokens": 32401568.0,
      "step": 18680
    },
    {
      "entropy": 5.578578758239746,
      "epoch": 1.4537638591713673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789827176979022,
      "loss": 5.0278,
      "mean_token_accuracy": 0.19821786135435104,
      "num_tokens": 32409996.0,
      "step": 18685
    },
    {
      "entropy": 5.562967300415039,
      "epoch": 1.454152888543085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004789710198166864,
      "loss": 5.0791,
      "mean_token_accuracy": 0.19199377596378325,
      "num_tokens": 32418886.0,
      "step": 18690
    },
    {
      "entropy": 5.554998779296875,
      "epoch": 1.4545419179148027,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047895931884053497,
      "loss": 4.9199,
      "mean_token_accuracy": 0.20571722090244293,
      "num_tokens": 32427999.0,
      "step": 18695
    },
    {
      "entropy": 5.645020246505737,
      "epoch": 1.4549309472865202,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047894761476962547,
      "loss": 5.1948,
      "mean_token_accuracy": 0.18876074105501175,
      "num_tokens": 32437115.0,
      "step": 18700
    },
    {
      "entropy": 5.613612699508667,
      "epoch": 1.4553199766582376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047893590760413545,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19629839658737183,
      "num_tokens": 32445668.0,
      "step": 18705
    },
    {
      "entropy": 5.568839025497437,
      "epoch": 1.4557090060299553,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047892419734424276,
      "loss": 5.0292,
      "mean_token_accuracy": 0.19784554839134216,
      "num_tokens": 32454022.0,
      "step": 18710
    },
    {
      "entropy": 5.684942102432251,
      "epoch": 1.4560980354016728,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047891248399012495,
      "loss": 5.1581,
      "mean_token_accuracy": 0.19272076487541198,
      "num_tokens": 32462997.0,
      "step": 18715
    },
    {
      "entropy": 5.672219276428223,
      "epoch": 1.4564870647733903,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004789007675419597,
      "loss": 5.0554,
      "mean_token_accuracy": 0.19555702209472656,
      "num_tokens": 32471184.0,
      "step": 18720
    },
    {
      "entropy": 5.595153570175171,
      "epoch": 1.456876094145108,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047888904799992486,
      "loss": 5.0701,
      "mean_token_accuracy": 0.19524910748004914,
      "num_tokens": 32480051.0,
      "step": 18725
    },
    {
      "entropy": 5.600559425354004,
      "epoch": 1.4572651235168255,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047887732536419826,
      "loss": 5.0892,
      "mean_token_accuracy": 0.19392136633396148,
      "num_tokens": 32489269.0,
      "step": 18730
    },
    {
      "entropy": 5.610526990890503,
      "epoch": 1.4576541528885432,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004788655996349577,
      "loss": 5.0513,
      "mean_token_accuracy": 0.1965191125869751,
      "num_tokens": 32497941.0,
      "step": 18735
    },
    {
      "entropy": 5.631112194061279,
      "epoch": 1.4580431822602606,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047885387081238125,
      "loss": 5.083,
      "mean_token_accuracy": 0.19369185119867324,
      "num_tokens": 32507679.0,
      "step": 18740
    },
    {
      "entropy": 5.601894903182983,
      "epoch": 1.4584322116319783,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004788421388966467,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19012330919504167,
      "num_tokens": 32516964.0,
      "step": 18745
    },
    {
      "entropy": 5.617612743377686,
      "epoch": 1.4588212410036958,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047883040388793207,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2072213277220726,
      "num_tokens": 32525178.0,
      "step": 18750
    },
    {
      "entropy": 5.670343923568725,
      "epoch": 1.4592102703754133,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047881866578641563,
      "loss": 5.1186,
      "mean_token_accuracy": 0.19039619266986846,
      "num_tokens": 32533694.0,
      "step": 18755
    },
    {
      "entropy": 5.565167188644409,
      "epoch": 1.459599299747131,
      "grad_norm": 1.125,
      "learning_rate": 0.0004788069245922753,
      "loss": 5.1005,
      "mean_token_accuracy": 0.1929655507206917,
      "num_tokens": 32542529.0,
      "step": 18760
    },
    {
      "entropy": 5.623377466201783,
      "epoch": 1.4599883291188485,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004787951803056893,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19313514828681946,
      "num_tokens": 32550987.0,
      "step": 18765
    },
    {
      "entropy": 5.564407157897949,
      "epoch": 1.460377358490566,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004787834329268358,
      "loss": 5.033,
      "mean_token_accuracy": 0.20288823395967484,
      "num_tokens": 32559812.0,
      "step": 18770
    },
    {
      "entropy": 5.579771661758423,
      "epoch": 1.4607663878622836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000478771682455893,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20041427314281463,
      "num_tokens": 32568477.0,
      "step": 18775
    },
    {
      "entropy": 5.722165107727051,
      "epoch": 1.461155417234001,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004787599288930393,
      "loss": 5.1241,
      "mean_token_accuracy": 0.18869419246912003,
      "num_tokens": 32577921.0,
      "step": 18780
    },
    {
      "entropy": 5.626677656173706,
      "epoch": 1.4615444466057188,
      "grad_norm": 1.078125,
      "learning_rate": 0.000478748172238453,
      "loss": 5.0259,
      "mean_token_accuracy": 0.19656828790903091,
      "num_tokens": 32587028.0,
      "step": 18785
    },
    {
      "entropy": 5.570806837081909,
      "epoch": 1.4619334759774363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004787364124923125,
      "loss": 5.0754,
      "mean_token_accuracy": 0.19527378380298616,
      "num_tokens": 32596092.0,
      "step": 18790
    },
    {
      "entropy": 5.630416536331177,
      "epoch": 1.462322505349154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047872464965479625,
      "loss": 5.014,
      "mean_token_accuracy": 0.1955581411719322,
      "num_tokens": 32605271.0,
      "step": 18795
    },
    {
      "entropy": 5.671950483322144,
      "epoch": 1.4627115347208715,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004787128837260826,
      "loss": 5.1951,
      "mean_token_accuracy": 0.18913070112466812,
      "num_tokens": 32614240.0,
      "step": 18800
    },
    {
      "entropy": 5.639273023605346,
      "epoch": 1.463100564092589,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004787011147063503,
      "loss": 5.1101,
      "mean_token_accuracy": 0.1965263843536377,
      "num_tokens": 32623390.0,
      "step": 18805
    },
    {
      "entropy": 5.613516187667846,
      "epoch": 1.4634895934643066,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004786893425957777,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19547268599271775,
      "num_tokens": 32631649.0,
      "step": 18810
    },
    {
      "entropy": 5.532551574707031,
      "epoch": 1.463878622836024,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004786775673945436,
      "loss": 5.0099,
      "mean_token_accuracy": 0.1957130953669548,
      "num_tokens": 32640844.0,
      "step": 18815
    },
    {
      "entropy": 5.629459953308105,
      "epoch": 1.4642676522077416,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047866578910282656,
      "loss": 5.0883,
      "mean_token_accuracy": 0.1940128982067108,
      "num_tokens": 32649458.0,
      "step": 18820
    },
    {
      "entropy": 5.588356351852417,
      "epoch": 1.4646566815794593,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047865400772080535,
      "loss": 5.0731,
      "mean_token_accuracy": 0.19747387021780013,
      "num_tokens": 32657649.0,
      "step": 18825
    },
    {
      "entropy": 5.663651180267334,
      "epoch": 1.4650457109511767,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047864222324865875,
      "loss": 5.1958,
      "mean_token_accuracy": 0.1885716736316681,
      "num_tokens": 32665407.0,
      "step": 18830
    },
    {
      "entropy": 5.622545766830444,
      "epoch": 1.4654347403228944,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004786304356865655,
      "loss": 5.051,
      "mean_token_accuracy": 0.1942585751414299,
      "num_tokens": 32673760.0,
      "step": 18835
    },
    {
      "entropy": 5.586823129653931,
      "epoch": 1.465823769694612,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047861864503470457,
      "loss": 5.0841,
      "mean_token_accuracy": 0.18849880993366241,
      "num_tokens": 32682118.0,
      "step": 18840
    },
    {
      "entropy": 5.594356727600098,
      "epoch": 1.4662127990663296,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004786068512932547,
      "loss": 5.092,
      "mean_token_accuracy": 0.19503487795591354,
      "num_tokens": 32690430.0,
      "step": 18845
    },
    {
      "entropy": 5.5979735374450685,
      "epoch": 1.466601828438047,
      "grad_norm": 1.15625,
      "learning_rate": 0.000478595054462395,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20188041925430297,
      "num_tokens": 32698776.0,
      "step": 18850
    },
    {
      "entropy": 5.6843665599822994,
      "epoch": 1.4669908578097646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047858325454230437,
      "loss": 5.1635,
      "mean_token_accuracy": 0.18863103687763214,
      "num_tokens": 32707299.0,
      "step": 18855
    },
    {
      "entropy": 5.560207843780518,
      "epoch": 1.4673798871814823,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004785714515331619,
      "loss": 5.0112,
      "mean_token_accuracy": 0.19962745904922485,
      "num_tokens": 32716041.0,
      "step": 18860
    },
    {
      "entropy": 5.650278615951538,
      "epoch": 1.4677689165531997,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047855964543514666,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20126706063747407,
      "num_tokens": 32724182.0,
      "step": 18865
    },
    {
      "entropy": 5.622026443481445,
      "epoch": 1.4681579459249172,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047854783624843786,
      "loss": 5.0064,
      "mean_token_accuracy": 0.19762877076864244,
      "num_tokens": 32732858.0,
      "step": 18870
    },
    {
      "entropy": 5.6472372055053714,
      "epoch": 1.468546975296635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047853602397321453,
      "loss": 5.1142,
      "mean_token_accuracy": 0.19586739987134932,
      "num_tokens": 32741422.0,
      "step": 18875
    },
    {
      "entropy": 5.6845897197723385,
      "epoch": 1.4689360046683524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047852420860965605,
      "loss": 5.1826,
      "mean_token_accuracy": 0.18357041627168655,
      "num_tokens": 32750720.0,
      "step": 18880
    },
    {
      "entropy": 5.609658241271973,
      "epoch": 1.46932503404007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047851239015794166,
      "loss": 5.0316,
      "mean_token_accuracy": 0.1995936304330826,
      "num_tokens": 32759418.0,
      "step": 18885
    },
    {
      "entropy": 5.6027069091796875,
      "epoch": 1.4697140634117876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047850056861825066,
      "loss": 5.0819,
      "mean_token_accuracy": 0.19997840672731398,
      "num_tokens": 32767604.0,
      "step": 18890
    },
    {
      "entropy": 5.603042316436768,
      "epoch": 1.4701030927835053,
      "grad_norm": 1.125,
      "learning_rate": 0.00047848874399076245,
      "loss": 5.1738,
      "mean_token_accuracy": 0.18337154388427734,
      "num_tokens": 32776535.0,
      "step": 18895
    },
    {
      "entropy": 5.695736455917358,
      "epoch": 1.4704921221552227,
      "grad_norm": 1.125,
      "learning_rate": 0.0004784769162756563,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19465018510818483,
      "num_tokens": 32785633.0,
      "step": 18900
    },
    {
      "entropy": 5.656501960754395,
      "epoch": 1.4708811515269402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000478465085473112,
      "loss": 5.0376,
      "mean_token_accuracy": 0.19833078235387802,
      "num_tokens": 32794281.0,
      "step": 18905
    },
    {
      "entropy": 5.674152135848999,
      "epoch": 1.471270180898658,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004784532515833088,
      "loss": 5.2095,
      "mean_token_accuracy": 0.19003974348306657,
      "num_tokens": 32802312.0,
      "step": 18910
    },
    {
      "entropy": 5.550491571426392,
      "epoch": 1.4716592102703754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047844141460642636,
      "loss": 4.99,
      "mean_token_accuracy": 0.19968778192996978,
      "num_tokens": 32810983.0,
      "step": 18915
    },
    {
      "entropy": 5.599526453018188,
      "epoch": 1.4720482396420929,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047842957454264425,
      "loss": 5.0281,
      "mean_token_accuracy": 0.19754729866981507,
      "num_tokens": 32820695.0,
      "step": 18920
    },
    {
      "entropy": 5.650800514221191,
      "epoch": 1.4724372690138106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047841773139214213,
      "loss": 5.0859,
      "mean_token_accuracy": 0.20043248683214188,
      "num_tokens": 32828483.0,
      "step": 18925
    },
    {
      "entropy": 5.597155570983887,
      "epoch": 1.472826298385528,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004784058851550997,
      "loss": 5.1008,
      "mean_token_accuracy": 0.1919277161359787,
      "num_tokens": 32837827.0,
      "step": 18930
    },
    {
      "entropy": 5.65845136642456,
      "epoch": 1.4732153277572457,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004783940358316967,
      "loss": 5.186,
      "mean_token_accuracy": 0.19253099113702773,
      "num_tokens": 32846872.0,
      "step": 18935
    },
    {
      "entropy": 5.604897356033325,
      "epoch": 1.4736043571289632,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047838218342211296,
      "loss": 4.996,
      "mean_token_accuracy": 0.2036880671977997,
      "num_tokens": 32855418.0,
      "step": 18940
    },
    {
      "entropy": 5.609240770339966,
      "epoch": 1.473993386500681,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047837032792652837,
      "loss": 5.0663,
      "mean_token_accuracy": 0.18735317885875702,
      "num_tokens": 32864075.0,
      "step": 18945
    },
    {
      "entropy": 5.709239101409912,
      "epoch": 1.4743824158723984,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004783584693451226,
      "loss": 5.1671,
      "mean_token_accuracy": 0.1834338590502739,
      "num_tokens": 32872615.0,
      "step": 18950
    },
    {
      "entropy": 5.6514629364013675,
      "epoch": 1.4747714452441159,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004783466076780759,
      "loss": 5.0222,
      "mean_token_accuracy": 0.1999700739979744,
      "num_tokens": 32881313.0,
      "step": 18955
    },
    {
      "entropy": 5.650282669067383,
      "epoch": 1.4751604746158336,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004783347429255679,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19320466816425325,
      "num_tokens": 32889323.0,
      "step": 18960
    },
    {
      "entropy": 5.525713872909546,
      "epoch": 1.475549503987551,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000478322875087779,
      "loss": 5.0621,
      "mean_token_accuracy": 0.18808478713035584,
      "num_tokens": 32898045.0,
      "step": 18965
    },
    {
      "entropy": 5.669640779495239,
      "epoch": 1.4759385333592685,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047831100416488906,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1874663531780243,
      "num_tokens": 32906646.0,
      "step": 18970
    },
    {
      "entropy": 5.717529201507569,
      "epoch": 1.4763275627309862,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047829913015707816,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19372713565826416,
      "num_tokens": 32915714.0,
      "step": 18975
    },
    {
      "entropy": 5.664337778091431,
      "epoch": 1.4767165921027037,
      "grad_norm": 1.125,
      "learning_rate": 0.00047828725306452657,
      "loss": 5.1828,
      "mean_token_accuracy": 0.1857110306620598,
      "num_tokens": 32924996.0,
      "step": 18980
    },
    {
      "entropy": 5.597547101974487,
      "epoch": 1.4771056214744214,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047827537288741453,
      "loss": 5.0102,
      "mean_token_accuracy": 0.19503621757030487,
      "num_tokens": 32933550.0,
      "step": 18985
    },
    {
      "entropy": 5.697243976593017,
      "epoch": 1.4774946508461388,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004782634896259222,
      "loss": 5.152,
      "mean_token_accuracy": 0.19050627052783967,
      "num_tokens": 32941800.0,
      "step": 18990
    },
    {
      "entropy": 5.63497405052185,
      "epoch": 1.4778836802178565,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047825160328023,
      "loss": 5.1127,
      "mean_token_accuracy": 0.19708827883005142,
      "num_tokens": 32951674.0,
      "step": 18995
    },
    {
      "entropy": 5.607395362854004,
      "epoch": 1.478272709589574,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004782397138505181,
      "loss": 5.047,
      "mean_token_accuracy": 0.19643496721982956,
      "num_tokens": 32959880.0,
      "step": 19000
    },
    {
      "entropy": 5.575069713592529,
      "epoch": 1.4786617389612915,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047822782133696715,
      "loss": 5.027,
      "mean_token_accuracy": 0.1995047315955162,
      "num_tokens": 32968557.0,
      "step": 19005
    },
    {
      "entropy": 5.743048191070557,
      "epoch": 1.4790507683330092,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004782159257397574,
      "loss": 5.1486,
      "mean_token_accuracy": 0.18862138986587523,
      "num_tokens": 32977519.0,
      "step": 19010
    },
    {
      "entropy": 5.725343227386475,
      "epoch": 1.4794397977047267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047820402705906953,
      "loss": 5.1376,
      "mean_token_accuracy": 0.19108114540576934,
      "num_tokens": 32985917.0,
      "step": 19015
    },
    {
      "entropy": 5.604153203964233,
      "epoch": 1.4798288270764441,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047819212529508394,
      "loss": 5.0512,
      "mean_token_accuracy": 0.1933942511677742,
      "num_tokens": 32994612.0,
      "step": 19020
    },
    {
      "entropy": 5.633587408065796,
      "epoch": 1.4802178564481618,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004781802204479812,
      "loss": 5.1626,
      "mean_token_accuracy": 0.19048628807067872,
      "num_tokens": 33003528.0,
      "step": 19025
    },
    {
      "entropy": 5.592432117462158,
      "epoch": 1.4806068858198795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004781683125179421,
      "loss": 5.0436,
      "mean_token_accuracy": 0.19696327447891235,
      "num_tokens": 33011685.0,
      "step": 19030
    },
    {
      "entropy": 5.632379341125488,
      "epoch": 1.480995915191597,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004781564015051472,
      "loss": 5.0796,
      "mean_token_accuracy": 0.19943165630102158,
      "num_tokens": 33020313.0,
      "step": 19035
    },
    {
      "entropy": 5.614678573608399,
      "epoch": 1.4813849445633145,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004781444874097772,
      "loss": 5.1136,
      "mean_token_accuracy": 0.19318949580192565,
      "num_tokens": 33029143.0,
      "step": 19040
    },
    {
      "entropy": 5.621893692016601,
      "epoch": 1.4817739739350322,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047813257023201307,
      "loss": 5.0499,
      "mean_token_accuracy": 0.19164162874221802,
      "num_tokens": 33038359.0,
      "step": 19045
    },
    {
      "entropy": 5.541339445114136,
      "epoch": 1.4821630033067497,
      "grad_norm": 1.125,
      "learning_rate": 0.0004781206499720354,
      "loss": 5.0978,
      "mean_token_accuracy": 0.1905768021941185,
      "num_tokens": 33046914.0,
      "step": 19050
    },
    {
      "entropy": 5.711096858978271,
      "epoch": 1.4825520326784671,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047810872663002523,
      "loss": 5.1683,
      "mean_token_accuracy": 0.18687887340784073,
      "num_tokens": 33054490.0,
      "step": 19055
    },
    {
      "entropy": 5.700817632675171,
      "epoch": 1.4829410620501848,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047809680020616333,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18899137228727342,
      "num_tokens": 33064624.0,
      "step": 19060
    },
    {
      "entropy": 5.577709531784057,
      "epoch": 1.4833300914219023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047808487070063083,
      "loss": 4.954,
      "mean_token_accuracy": 0.19713495969772338,
      "num_tokens": 33072581.0,
      "step": 19065
    },
    {
      "entropy": 5.645633697509766,
      "epoch": 1.4837191207936198,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004780729381136086,
      "loss": 5.0939,
      "mean_token_accuracy": 0.19245125949382783,
      "num_tokens": 33081696.0,
      "step": 19070
    },
    {
      "entropy": 5.658846807479859,
      "epoch": 1.4841081501653375,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004780610024452778,
      "loss": 5.1063,
      "mean_token_accuracy": 0.18965048491954803,
      "num_tokens": 33090470.0,
      "step": 19075
    },
    {
      "entropy": 5.69404091835022,
      "epoch": 1.4844971795370552,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047804906369581954,
      "loss": 5.1461,
      "mean_token_accuracy": 0.19724612236022948,
      "num_tokens": 33099006.0,
      "step": 19080
    },
    {
      "entropy": 5.577608871459961,
      "epoch": 1.4848862089087727,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004780371218654148,
      "loss": 4.9819,
      "mean_token_accuracy": 0.20557599812746047,
      "num_tokens": 33107715.0,
      "step": 19085
    },
    {
      "entropy": 5.614643716812134,
      "epoch": 1.4852752382804901,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047802517695424496,
      "loss": 5.0415,
      "mean_token_accuracy": 0.20244930684566498,
      "num_tokens": 33116895.0,
      "step": 19090
    },
    {
      "entropy": 5.6015589237213135,
      "epoch": 1.4856642676522078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004780132289624913,
      "loss": 5.0819,
      "mean_token_accuracy": 0.19595806002616883,
      "num_tokens": 33125676.0,
      "step": 19095
    },
    {
      "entropy": 5.617783164978027,
      "epoch": 1.4860532970239253,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004780012778903349,
      "loss": 5.0794,
      "mean_token_accuracy": 0.19584878534078598,
      "num_tokens": 33134870.0,
      "step": 19100
    },
    {
      "entropy": 5.644699239730835,
      "epoch": 1.4864423263956428,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047798932373795724,
      "loss": 5.1367,
      "mean_token_accuracy": 0.1970182940363884,
      "num_tokens": 33144258.0,
      "step": 19105
    },
    {
      "entropy": 5.650774908065796,
      "epoch": 1.4868313557673605,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047797736650553977,
      "loss": 5.0911,
      "mean_token_accuracy": 0.1943137139081955,
      "num_tokens": 33153703.0,
      "step": 19110
    },
    {
      "entropy": 5.647247982025147,
      "epoch": 1.487220385139078,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047796540619326366,
      "loss": 5.0965,
      "mean_token_accuracy": 0.18595061600208282,
      "num_tokens": 33163092.0,
      "step": 19115
    },
    {
      "entropy": 5.616321134567261,
      "epoch": 1.4876094145107956,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004779534428013107,
      "loss": 5.1033,
      "mean_token_accuracy": 0.19394506514072418,
      "num_tokens": 33172133.0,
      "step": 19120
    },
    {
      "entropy": 5.602980327606201,
      "epoch": 1.4879984438825131,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047794147632986223,
      "loss": 5.11,
      "mean_token_accuracy": 0.18816616982221604,
      "num_tokens": 33180629.0,
      "step": 19125
    },
    {
      "entropy": 5.600535345077515,
      "epoch": 1.4883874732542308,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004779295067790999,
      "loss": 5.0411,
      "mean_token_accuracy": 0.19589447975158691,
      "num_tokens": 33189602.0,
      "step": 19130
    },
    {
      "entropy": 5.660037612915039,
      "epoch": 1.4887765026259483,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004779175341492053,
      "loss": 5.096,
      "mean_token_accuracy": 0.1949512094259262,
      "num_tokens": 33198403.0,
      "step": 19135
    },
    {
      "entropy": 5.5873847007751465,
      "epoch": 1.4891655319976658,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004779055584403601,
      "loss": 5.0354,
      "mean_token_accuracy": 0.19470856487751007,
      "num_tokens": 33207131.0,
      "step": 19140
    },
    {
      "entropy": 5.611118030548096,
      "epoch": 1.4895545613693835,
      "grad_norm": 1.109375,
      "learning_rate": 0.000477893579652746,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19131521582603456,
      "num_tokens": 33216143.0,
      "step": 19145
    },
    {
      "entropy": 5.605795907974243,
      "epoch": 1.489943590741101,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047788159778654475,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20035088211297988,
      "num_tokens": 33224558.0,
      "step": 19150
    },
    {
      "entropy": 5.523874664306641,
      "epoch": 1.4903326201128184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047786961284193813,
      "loss": 5.0786,
      "mean_token_accuracy": 0.19427258521318436,
      "num_tokens": 33232737.0,
      "step": 19155
    },
    {
      "entropy": 5.621190404891967,
      "epoch": 1.4907216494845361,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004778576248191081,
      "loss": 5.0719,
      "mean_token_accuracy": 0.195353165268898,
      "num_tokens": 33240845.0,
      "step": 19160
    },
    {
      "entropy": 5.661509275436401,
      "epoch": 1.4911106788562536,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004778456337182365,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20308296233415604,
      "num_tokens": 33249101.0,
      "step": 19165
    },
    {
      "entropy": 5.678815460205078,
      "epoch": 1.4914997082279713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004778336395395052,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19764333516359328,
      "num_tokens": 33258403.0,
      "step": 19170
    },
    {
      "entropy": 5.553795862197876,
      "epoch": 1.4918887375996888,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047782164228309636,
      "loss": 5.0043,
      "mean_token_accuracy": 0.1974925622344017,
      "num_tokens": 33266679.0,
      "step": 19175
    },
    {
      "entropy": 5.570895433425903,
      "epoch": 1.4922777669714065,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047780964194919193,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20399714559316634,
      "num_tokens": 33274996.0,
      "step": 19180
    },
    {
      "entropy": 5.711767530441284,
      "epoch": 1.492666796343124,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047779763853797387,
      "loss": 5.1497,
      "mean_token_accuracy": 0.1916724696755409,
      "num_tokens": 33282558.0,
      "step": 19185
    },
    {
      "entropy": 5.628012943267822,
      "epoch": 1.4930558257148414,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004777856320496245,
      "loss": 5.1229,
      "mean_token_accuracy": 0.18385151773691177,
      "num_tokens": 33291233.0,
      "step": 19190
    },
    {
      "entropy": 5.6110797882080075,
      "epoch": 1.493444855086559,
      "grad_norm": 1.125,
      "learning_rate": 0.000477773622484326,
      "loss": 4.9879,
      "mean_token_accuracy": 0.19824894070625304,
      "num_tokens": 33299821.0,
      "step": 19195
    },
    {
      "entropy": 5.603002834320068,
      "epoch": 1.4938338844582766,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004777616098422604,
      "loss": 5.0837,
      "mean_token_accuracy": 0.1970875307917595,
      "num_tokens": 33309624.0,
      "step": 19200
    },
    {
      "entropy": 5.622921371459961,
      "epoch": 1.494222913829994,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047774959412361014,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19579890817403794,
      "num_tokens": 33318727.0,
      "step": 19205
    },
    {
      "entropy": 5.7048276424407955,
      "epoch": 1.4946119432017118,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004777375753285575,
      "loss": 5.0898,
      "mean_token_accuracy": 0.1956559717655182,
      "num_tokens": 33327219.0,
      "step": 19210
    },
    {
      "entropy": 5.609356355667114,
      "epoch": 1.4950009725734292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047772555345728486,
      "loss": 5.1007,
      "mean_token_accuracy": 0.19869064092636107,
      "num_tokens": 33336466.0,
      "step": 19215
    },
    {
      "entropy": 5.7098297595977785,
      "epoch": 1.495390001945147,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004777135285099746,
      "loss": 5.1524,
      "mean_token_accuracy": 0.1837605983018875,
      "num_tokens": 33346040.0,
      "step": 19220
    },
    {
      "entropy": 5.723146677017212,
      "epoch": 1.4957790313168644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004777015004868092,
      "loss": 5.0897,
      "mean_token_accuracy": 0.1936156451702118,
      "num_tokens": 33355476.0,
      "step": 19225
    },
    {
      "entropy": 5.605170249938965,
      "epoch": 1.496168060688582,
      "grad_norm": 1.125,
      "learning_rate": 0.0004776894693879712,
      "loss": 5.0646,
      "mean_token_accuracy": 0.2019073784351349,
      "num_tokens": 33364455.0,
      "step": 19230
    },
    {
      "entropy": 5.550471496582031,
      "epoch": 1.4965570900602996,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004776774352136431,
      "loss": 5.0429,
      "mean_token_accuracy": 0.19171344190835954,
      "num_tokens": 33373129.0,
      "step": 19235
    },
    {
      "entropy": 5.590015172958374,
      "epoch": 1.496946119432017,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047766539796400756,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1957753598690033,
      "num_tokens": 33382775.0,
      "step": 19240
    },
    {
      "entropy": 5.620344829559326,
      "epoch": 1.4973351488037348,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004776533576392471,
      "loss": 5.0807,
      "mean_token_accuracy": 0.19079612344503402,
      "num_tokens": 33391370.0,
      "step": 19245
    },
    {
      "entropy": 5.700890350341797,
      "epoch": 1.4977241781754522,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004776413142395445,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20143744200468064,
      "num_tokens": 33399832.0,
      "step": 19250
    },
    {
      "entropy": 5.629574537277222,
      "epoch": 1.4981132075471697,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047762926776508244,
      "loss": 5.1049,
      "mean_token_accuracy": 0.1914284661412239,
      "num_tokens": 33408709.0,
      "step": 19255
    },
    {
      "entropy": 5.61508092880249,
      "epoch": 1.4985022369188874,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047761721821604387,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1863863378763199,
      "num_tokens": 33416983.0,
      "step": 19260
    },
    {
      "entropy": 5.644004678726196,
      "epoch": 1.4988912662906049,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004776051655926115,
      "loss": 5.1154,
      "mean_token_accuracy": 0.19863494336605073,
      "num_tokens": 33426692.0,
      "step": 19265
    },
    {
      "entropy": 5.667902946472168,
      "epoch": 1.4992802956623226,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047759310989496813,
      "loss": 5.1057,
      "mean_token_accuracy": 0.2003609448671341,
      "num_tokens": 33434895.0,
      "step": 19270
    },
    {
      "entropy": 5.5761466979980465,
      "epoch": 1.49966932503404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004775810511232969,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19568276405334473,
      "num_tokens": 33443378.0,
      "step": 19275
    },
    {
      "entropy": 5.671515083312988,
      "epoch": 1.5000583544057577,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047756898927778056,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19509005695581436,
      "num_tokens": 33452753.0,
      "step": 19280
    },
    {
      "entropy": 5.743964004516601,
      "epoch": 1.5004473837774752,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047755692435860227,
      "loss": 5.1205,
      "mean_token_accuracy": 0.1881059244275093,
      "num_tokens": 33461625.0,
      "step": 19285
    },
    {
      "entropy": 5.616452741622925,
      "epoch": 1.5008364131491927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000477544856365945,
      "loss": 5.0512,
      "mean_token_accuracy": 0.19807957112789154,
      "num_tokens": 33471228.0,
      "step": 19290
    },
    {
      "entropy": 5.599948167800903,
      "epoch": 1.5012254425209104,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047753278529999205,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20140955746173858,
      "num_tokens": 33480137.0,
      "step": 19295
    },
    {
      "entropy": 5.6027764797210695,
      "epoch": 1.5016144718926279,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047752071116092637,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20162603110074998,
      "num_tokens": 33488860.0,
      "step": 19300
    },
    {
      "entropy": 5.642009305953979,
      "epoch": 1.5020035012643453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004775086339489312,
      "loss": 5.0359,
      "mean_token_accuracy": 0.19391362816095353,
      "num_tokens": 33497625.0,
      "step": 19305
    },
    {
      "entropy": 5.651634359359742,
      "epoch": 1.502392530636063,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004774965536641899,
      "loss": 5.1969,
      "mean_token_accuracy": 0.18374914228916167,
      "num_tokens": 33506422.0,
      "step": 19310
    },
    {
      "entropy": 5.559615421295166,
      "epoch": 1.5027815600077807,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047748447030688575,
      "loss": 4.9881,
      "mean_token_accuracy": 0.19838032573461534,
      "num_tokens": 33515638.0,
      "step": 19315
    },
    {
      "entropy": 5.647777318954468,
      "epoch": 1.503170589379498,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047747238387720194,
      "loss": 5.0108,
      "mean_token_accuracy": 0.1923777639865875,
      "num_tokens": 33524287.0,
      "step": 19320
    },
    {
      "entropy": 5.628607749938965,
      "epoch": 1.5035596187512157,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000477460294375322,
      "loss": 5.0175,
      "mean_token_accuracy": 0.19014522582292556,
      "num_tokens": 33532579.0,
      "step": 19325
    },
    {
      "entropy": 5.56951937675476,
      "epoch": 1.5039486481229334,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047744820180142935,
      "loss": 5.1008,
      "mean_token_accuracy": 0.19044699966907502,
      "num_tokens": 33540816.0,
      "step": 19330
    },
    {
      "entropy": 5.526912355422974,
      "epoch": 1.5043376774946509,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047743610615570746,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20500442236661912,
      "num_tokens": 33549383.0,
      "step": 19335
    },
    {
      "entropy": 5.53729944229126,
      "epoch": 1.5047267068663683,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047742400743833993,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20245888978242874,
      "num_tokens": 33557457.0,
      "step": 19340
    },
    {
      "entropy": 5.658443212509155,
      "epoch": 1.505115736238086,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004774119056495102,
      "loss": 5.1164,
      "mean_token_accuracy": 0.1955449879169464,
      "num_tokens": 33565479.0,
      "step": 19345
    },
    {
      "entropy": 5.629749727249146,
      "epoch": 1.5055047656098035,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047739980078940206,
      "loss": 5.0589,
      "mean_token_accuracy": 0.19207566678524018,
      "num_tokens": 33574691.0,
      "step": 19350
    },
    {
      "entropy": 5.552569580078125,
      "epoch": 1.505893794981521,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047738769285819894,
      "loss": 4.931,
      "mean_token_accuracy": 0.20868074595928193,
      "num_tokens": 33582360.0,
      "step": 19355
    },
    {
      "entropy": 5.5539939403533936,
      "epoch": 1.5062828243532387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004773755818560849,
      "loss": 5.0768,
      "mean_token_accuracy": 0.19583491533994674,
      "num_tokens": 33590977.0,
      "step": 19360
    },
    {
      "entropy": 5.5635261058807375,
      "epoch": 1.5066718537249564,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004773634677832434,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19161334931850432,
      "num_tokens": 33600227.0,
      "step": 19365
    },
    {
      "entropy": 5.666249322891235,
      "epoch": 1.5070608830966739,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004773513506398584,
      "loss": 5.072,
      "mean_token_accuracy": 0.20054676085710527,
      "num_tokens": 33608408.0,
      "step": 19370
    },
    {
      "entropy": 5.66262845993042,
      "epoch": 1.5074499124683913,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004773392304261137,
      "loss": 5.0131,
      "mean_token_accuracy": 0.19590090662240983,
      "num_tokens": 33616892.0,
      "step": 19375
    },
    {
      "entropy": 5.632138586044311,
      "epoch": 1.507838941840109,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004773271071421933,
      "loss": 5.0972,
      "mean_token_accuracy": 0.19390486180782318,
      "num_tokens": 33626090.0,
      "step": 19380
    },
    {
      "entropy": 5.5055474758148195,
      "epoch": 1.5082279712118265,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047731498078828105,
      "loss": 4.9639,
      "mean_token_accuracy": 0.19982597082853318,
      "num_tokens": 33634855.0,
      "step": 19385
    },
    {
      "entropy": 5.6039186954498295,
      "epoch": 1.508617000583544,
      "grad_norm": 1.171875,
      "learning_rate": 0.000477302851364561,
      "loss": 5.1075,
      "mean_token_accuracy": 0.18991513550281525,
      "num_tokens": 33643425.0,
      "step": 19390
    },
    {
      "entropy": 5.753226375579834,
      "epoch": 1.5090060299552617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047729071887121717,
      "loss": 5.2165,
      "mean_token_accuracy": 0.1843421682715416,
      "num_tokens": 33652926.0,
      "step": 19395
    },
    {
      "entropy": 5.736544036865235,
      "epoch": 1.5093950593269791,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004772785833084337,
      "loss": 5.1115,
      "mean_token_accuracy": 0.19314922392368317,
      "num_tokens": 33661279.0,
      "step": 19400
    },
    {
      "entropy": 5.614797639846802,
      "epoch": 1.5097840886986966,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004772664446763946,
      "loss": 5.0045,
      "mean_token_accuracy": 0.18962804973125458,
      "num_tokens": 33669717.0,
      "step": 19405
    },
    {
      "entropy": 5.649522972106934,
      "epoch": 1.5101731180704143,
      "grad_norm": 1.125,
      "learning_rate": 0.0004772543029752842,
      "loss": 5.1397,
      "mean_token_accuracy": 0.189462374150753,
      "num_tokens": 33679141.0,
      "step": 19410
    },
    {
      "entropy": 5.6186011791229244,
      "epoch": 1.510562147442132,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047724215820528666,
      "loss": 4.9484,
      "mean_token_accuracy": 0.19894350022077562,
      "num_tokens": 33686867.0,
      "step": 19415
    },
    {
      "entropy": 5.606513547897339,
      "epoch": 1.5109511768138495,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004772300103665863,
      "loss": 5.0477,
      "mean_token_accuracy": 0.20092494934797286,
      "num_tokens": 33695073.0,
      "step": 19420
    },
    {
      "entropy": 5.627023649215698,
      "epoch": 1.511340206185567,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047721785945936733,
      "loss": 5.142,
      "mean_token_accuracy": 0.18762203603982924,
      "num_tokens": 33703212.0,
      "step": 19425
    },
    {
      "entropy": 5.6336384296417235,
      "epoch": 1.5117292355572847,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004772057054838143,
      "loss": 5.0997,
      "mean_token_accuracy": 0.1948071837425232,
      "num_tokens": 33711487.0,
      "step": 19430
    },
    {
      "entropy": 5.670090007781982,
      "epoch": 1.5121182649290021,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047719354844011146,
      "loss": 5.1425,
      "mean_token_accuracy": 0.1909279316663742,
      "num_tokens": 33719642.0,
      "step": 19435
    },
    {
      "entropy": 5.610255241394043,
      "epoch": 1.5125072943007196,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004771813883284434,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2012290269136429,
      "num_tokens": 33728167.0,
      "step": 19440
    },
    {
      "entropy": 5.599192762374878,
      "epoch": 1.5128963236724373,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047716922514899455,
      "loss": 5.0825,
      "mean_token_accuracy": 0.2018330916762352,
      "num_tokens": 33736094.0,
      "step": 19445
    },
    {
      "entropy": 5.600958251953125,
      "epoch": 1.5132853530441548,
      "grad_norm": 1.125,
      "learning_rate": 0.00047715705890194953,
      "loss": 4.9525,
      "mean_token_accuracy": 0.20545677542686464,
      "num_tokens": 33744416.0,
      "step": 19450
    },
    {
      "entropy": 5.689314270019532,
      "epoch": 1.5136743824158723,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047714488958749285,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20169136971235274,
      "num_tokens": 33752873.0,
      "step": 19455
    },
    {
      "entropy": 5.587159204483032,
      "epoch": 1.51406341178759,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047713271720580924,
      "loss": 5.0321,
      "mean_token_accuracy": 0.1978490546345711,
      "num_tokens": 33761604.0,
      "step": 19460
    },
    {
      "entropy": 5.635813522338867,
      "epoch": 1.5144524411593077,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004771205417570834,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1977468177676201,
      "num_tokens": 33770562.0,
      "step": 19465
    },
    {
      "entropy": 5.63994026184082,
      "epoch": 1.5148414705310251,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047710836324150004,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20262469500303268,
      "num_tokens": 33779037.0,
      "step": 19470
    },
    {
      "entropy": 5.637336730957031,
      "epoch": 1.5152304999027426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000477096181659244,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19674476683139802,
      "num_tokens": 33787949.0,
      "step": 19475
    },
    {
      "entropy": 5.577021169662475,
      "epoch": 1.5156195292744603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004770839970105,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20090929120779039,
      "num_tokens": 33796783.0,
      "step": 19480
    },
    {
      "entropy": 5.620127153396607,
      "epoch": 1.5160085586461778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047707180929545295,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1963385045528412,
      "num_tokens": 33804975.0,
      "step": 19485
    },
    {
      "entropy": 5.577420997619629,
      "epoch": 1.5163975880178953,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047705961851428786,
      "loss": 5.0487,
      "mean_token_accuracy": 0.1972619667649269,
      "num_tokens": 33813355.0,
      "step": 19490
    },
    {
      "entropy": 5.593205833435059,
      "epoch": 1.516786617389613,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047704742466718973,
      "loss": 5.0189,
      "mean_token_accuracy": 0.20256625711917878,
      "num_tokens": 33821710.0,
      "step": 19495
    },
    {
      "entropy": 5.695212173461914,
      "epoch": 1.5171756467613304,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047703522775434354,
      "loss": 5.1648,
      "mean_token_accuracy": 0.18924276977777482,
      "num_tokens": 33830674.0,
      "step": 19500
    },
    {
      "entropy": 5.554477643966675,
      "epoch": 1.517564676133048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047702302777593425,
      "loss": 5.0084,
      "mean_token_accuracy": 0.19690649211406708,
      "num_tokens": 33838963.0,
      "step": 19505
    },
    {
      "entropy": 5.683591413497925,
      "epoch": 1.5179537055047656,
      "grad_norm": 1.125,
      "learning_rate": 0.00047701082473214715,
      "loss": 5.1481,
      "mean_token_accuracy": 0.1916452780365944,
      "num_tokens": 33847728.0,
      "step": 19510
    },
    {
      "entropy": 5.617996454238892,
      "epoch": 1.5183427348764833,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047699861862316725,
      "loss": 5.122,
      "mean_token_accuracy": 0.19821913093328475,
      "num_tokens": 33856170.0,
      "step": 19515
    },
    {
      "entropy": 5.6247189998626705,
      "epoch": 1.5187317642482008,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004769864094491798,
      "loss": 5.1032,
      "mean_token_accuracy": 0.19808651953935624,
      "num_tokens": 33865531.0,
      "step": 19520
    },
    {
      "entropy": 5.647815799713134,
      "epoch": 1.5191207936199183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004769741972103702,
      "loss": 5.0362,
      "mean_token_accuracy": 0.19149678200483322,
      "num_tokens": 33873976.0,
      "step": 19525
    },
    {
      "entropy": 5.589547348022461,
      "epoch": 1.519509822991636,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047696198190692353,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2095463365316391,
      "num_tokens": 33881884.0,
      "step": 19530
    },
    {
      "entropy": 5.59876594543457,
      "epoch": 1.5198988523633534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004769497635390253,
      "loss": 5.1304,
      "mean_token_accuracy": 0.19355648905038833,
      "num_tokens": 33890932.0,
      "step": 19535
    },
    {
      "entropy": 5.62343397140503,
      "epoch": 1.520287881735071,
      "grad_norm": 1.125,
      "learning_rate": 0.0004769375421068608,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19374216794967652,
      "num_tokens": 33899271.0,
      "step": 19540
    },
    {
      "entropy": 5.631417226791382,
      "epoch": 1.5206769111067886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047692531761061555,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19253962785005568,
      "num_tokens": 33908776.0,
      "step": 19545
    },
    {
      "entropy": 5.628013372421265,
      "epoch": 1.521065940478506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000476913090050475,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20432766824960708,
      "num_tokens": 33917424.0,
      "step": 19550
    },
    {
      "entropy": 5.690599679946899,
      "epoch": 1.5214549698502235,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047690085942662464,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1910681441426277,
      "num_tokens": 33925948.0,
      "step": 19555
    },
    {
      "entropy": 5.650610160827637,
      "epoch": 1.5218439992219412,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047688862573925015,
      "loss": 5.1032,
      "mean_token_accuracy": 0.19761233031749725,
      "num_tokens": 33934623.0,
      "step": 19560
    },
    {
      "entropy": 5.622818279266357,
      "epoch": 1.522233028593659,
      "grad_norm": 1.125,
      "learning_rate": 0.00047687638898853707,
      "loss": 5.1106,
      "mean_token_accuracy": 0.19777024686336517,
      "num_tokens": 33943871.0,
      "step": 19565
    },
    {
      "entropy": 5.645605230331421,
      "epoch": 1.5226220579653764,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004768641491746711,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19446071684360505,
      "num_tokens": 33953032.0,
      "step": 19570
    },
    {
      "entropy": 5.637818145751953,
      "epoch": 1.523011087337094,
      "grad_norm": 1.125,
      "learning_rate": 0.000476851906297838,
      "loss": 5.1075,
      "mean_token_accuracy": 0.19352239817380906,
      "num_tokens": 33961702.0,
      "step": 19575
    },
    {
      "entropy": 5.63962607383728,
      "epoch": 1.5234001167088116,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047683966035822346,
      "loss": 5.1334,
      "mean_token_accuracy": 0.18768450319767,
      "num_tokens": 33970996.0,
      "step": 19580
    },
    {
      "entropy": 5.659076309204101,
      "epoch": 1.523789146080529,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047682741135601336,
      "loss": 5.1014,
      "mean_token_accuracy": 0.1944100260734558,
      "num_tokens": 33980445.0,
      "step": 19585
    },
    {
      "entropy": 5.6907415866851805,
      "epoch": 1.5241781754522465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047681515929139356,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1970606654882431,
      "num_tokens": 33989042.0,
      "step": 19590
    },
    {
      "entropy": 5.546412754058838,
      "epoch": 1.5245672048239642,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047680290416454995,
      "loss": 4.9206,
      "mean_token_accuracy": 0.2149445965886116,
      "num_tokens": 33998162.0,
      "step": 19595
    },
    {
      "entropy": 5.603272247314453,
      "epoch": 1.524956234195682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004767906459756684,
      "loss": 5.0539,
      "mean_token_accuracy": 0.19917830377817153,
      "num_tokens": 34007480.0,
      "step": 19600
    },
    {
      "entropy": 5.660833311080933,
      "epoch": 1.5253452635673992,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047677838472493504,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19745680838823318,
      "num_tokens": 34015605.0,
      "step": 19605
    },
    {
      "entropy": 5.692202186584472,
      "epoch": 1.5257342929391169,
      "grad_norm": 1.125,
      "learning_rate": 0.0004767661204125358,
      "loss": 5.1679,
      "mean_token_accuracy": 0.1887834772467613,
      "num_tokens": 34024847.0,
      "step": 19610
    },
    {
      "entropy": 5.605807495117188,
      "epoch": 1.5261233223108346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004767538530386569,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1913454294204712,
      "num_tokens": 34034205.0,
      "step": 19615
    },
    {
      "entropy": 5.6310220718383786,
      "epoch": 1.526512351682552,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047674158260348437,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20102907866239547,
      "num_tokens": 34042578.0,
      "step": 19620
    },
    {
      "entropy": 5.599019384384155,
      "epoch": 1.5269013810542695,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047672930910720436,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19896105378866197,
      "num_tokens": 34051699.0,
      "step": 19625
    },
    {
      "entropy": 5.571763038635254,
      "epoch": 1.5272904104259872,
      "grad_norm": 1.125,
      "learning_rate": 0.00047671703255000326,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20116227567195893,
      "num_tokens": 34060872.0,
      "step": 19630
    },
    {
      "entropy": 5.647749996185302,
      "epoch": 1.5276794397977047,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004767047529320673,
      "loss": 5.0889,
      "mean_token_accuracy": 0.18749193102121353,
      "num_tokens": 34069501.0,
      "step": 19635
    },
    {
      "entropy": 5.666941165924072,
      "epoch": 1.5280684691694222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047669247025358257,
      "loss": 5.1329,
      "mean_token_accuracy": 0.1864587128162384,
      "num_tokens": 34078573.0,
      "step": 19640
    },
    {
      "entropy": 5.620309066772461,
      "epoch": 1.5284574985411399,
      "grad_norm": 1.125,
      "learning_rate": 0.00047668018451473584,
      "loss": 5.0981,
      "mean_token_accuracy": 0.19171562194824218,
      "num_tokens": 34087270.0,
      "step": 19645
    },
    {
      "entropy": 5.607879638671875,
      "epoch": 1.5288465279128576,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004766678957157132,
      "loss": 5.1273,
      "mean_token_accuracy": 0.1968986541032791,
      "num_tokens": 34096205.0,
      "step": 19650
    },
    {
      "entropy": 5.6425580978393555,
      "epoch": 1.5292355572845748,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004766556038567013,
      "loss": 4.96,
      "mean_token_accuracy": 0.20274903923273085,
      "num_tokens": 34104736.0,
      "step": 19655
    },
    {
      "entropy": 5.658192729949951,
      "epoch": 1.5296245866562925,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004766433089378865,
      "loss": 5.0508,
      "mean_token_accuracy": 0.19681010842323304,
      "num_tokens": 34113420.0,
      "step": 19660
    },
    {
      "entropy": 5.673188209533691,
      "epoch": 1.5300136160280102,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047663101095945544,
      "loss": 5.1707,
      "mean_token_accuracy": 0.18345843702554704,
      "num_tokens": 34121702.0,
      "step": 19665
    },
    {
      "entropy": 5.6602496147155765,
      "epoch": 1.5304026453997277,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047661870992159476,
      "loss": 5.0383,
      "mean_token_accuracy": 0.19873252660036086,
      "num_tokens": 34129657.0,
      "step": 19670
    },
    {
      "entropy": 5.671857643127441,
      "epoch": 1.5307916747714452,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047660640582449106,
      "loss": 5.1738,
      "mean_token_accuracy": 0.18919049948453903,
      "num_tokens": 34138078.0,
      "step": 19675
    },
    {
      "entropy": 5.66781907081604,
      "epoch": 1.5311807041431629,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047659409866833104,
      "loss": 5.069,
      "mean_token_accuracy": 0.1989251047372818,
      "num_tokens": 34146402.0,
      "step": 19680
    },
    {
      "entropy": 5.693409109115601,
      "epoch": 1.5315697335148803,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004765817884533014,
      "loss": 5.188,
      "mean_token_accuracy": 0.1895010605454445,
      "num_tokens": 34155909.0,
      "step": 19685
    },
    {
      "entropy": 5.589903402328491,
      "epoch": 1.5319587628865978,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000476569475179589,
      "loss": 5.0244,
      "mean_token_accuracy": 0.19652349799871444,
      "num_tokens": 34164598.0,
      "step": 19690
    },
    {
      "entropy": 5.6375257015228275,
      "epoch": 1.5323477922583155,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047655715884738074,
      "loss": 5.0822,
      "mean_token_accuracy": 0.19590800255537033,
      "num_tokens": 34173476.0,
      "step": 19695
    },
    {
      "entropy": 5.624957609176636,
      "epoch": 1.5327368216300332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004765448394568632,
      "loss": 5.1088,
      "mean_token_accuracy": 0.19063840359449385,
      "num_tokens": 34181986.0,
      "step": 19700
    },
    {
      "entropy": 5.643808889389038,
      "epoch": 1.5331258510017505,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004765325170082237,
      "loss": 5.0576,
      "mean_token_accuracy": 0.197055946290493,
      "num_tokens": 34190477.0,
      "step": 19705
    },
    {
      "entropy": 5.57037615776062,
      "epoch": 1.5335148803734682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000476520191501649,
      "loss": 5.0083,
      "mean_token_accuracy": 0.19944454580545426,
      "num_tokens": 34198955.0,
      "step": 19710
    },
    {
      "entropy": 5.643812036514282,
      "epoch": 1.5339039097451859,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047650786293732607,
      "loss": 5.1637,
      "mean_token_accuracy": 0.1872788667678833,
      "num_tokens": 34207744.0,
      "step": 19715
    },
    {
      "entropy": 5.64088282585144,
      "epoch": 1.5342929391169033,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004764955313154421,
      "loss": 5.0854,
      "mean_token_accuracy": 0.1977681264281273,
      "num_tokens": 34215935.0,
      "step": 19720
    },
    {
      "entropy": 5.597507953643799,
      "epoch": 1.5346819684886208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047648319663618415,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19041227400302888,
      "num_tokens": 34224990.0,
      "step": 19725
    },
    {
      "entropy": 5.594324111938477,
      "epoch": 1.5350709978603385,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047647085889973936,
      "loss": 5.0172,
      "mean_token_accuracy": 0.19416527599096298,
      "num_tokens": 34232928.0,
      "step": 19730
    },
    {
      "entropy": 5.583013248443604,
      "epoch": 1.535460027232056,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047645851810629503,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2047799974679947,
      "num_tokens": 34241434.0,
      "step": 19735
    },
    {
      "entropy": 5.550470018386841,
      "epoch": 1.5358490566037735,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047644617425603825,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19984989613294601,
      "num_tokens": 34250067.0,
      "step": 19740
    },
    {
      "entropy": 5.621737861633301,
      "epoch": 1.5362380859754912,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004764338273491565,
      "loss": 5.098,
      "mean_token_accuracy": 0.19577794820070266,
      "num_tokens": 34258656.0,
      "step": 19745
    },
    {
      "entropy": 5.6698202133178714,
      "epoch": 1.5366271153472089,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047642147738583695,
      "loss": 5.17,
      "mean_token_accuracy": 0.18066570907831192,
      "num_tokens": 34267417.0,
      "step": 19750
    },
    {
      "entropy": 5.627575492858886,
      "epoch": 1.5370161447189261,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004764091243662671,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21379761695861815,
      "num_tokens": 34275334.0,
      "step": 19755
    },
    {
      "entropy": 5.595056200027466,
      "epoch": 1.5374051740906438,
      "grad_norm": 1.25,
      "learning_rate": 0.00047639676829063437,
      "loss": 5.0999,
      "mean_token_accuracy": 0.19561252295970916,
      "num_tokens": 34284016.0,
      "step": 19760
    },
    {
      "entropy": 5.525123500823975,
      "epoch": 1.5377942034623615,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047638440915912623,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20568907260894775,
      "num_tokens": 34292747.0,
      "step": 19765
    },
    {
      "entropy": 5.5779155731201175,
      "epoch": 1.538183232834079,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004763720469719303,
      "loss": 4.9955,
      "mean_token_accuracy": 0.19860865026712418,
      "num_tokens": 34301764.0,
      "step": 19770
    },
    {
      "entropy": 5.639548969268799,
      "epoch": 1.5385722622057965,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000476359681729234,
      "loss": 5.0974,
      "mean_token_accuracy": 0.19451828747987748,
      "num_tokens": 34310522.0,
      "step": 19775
    },
    {
      "entropy": 5.695370674133301,
      "epoch": 1.5389612915775142,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000476347313431225,
      "loss": 5.0875,
      "mean_token_accuracy": 0.19985114485025407,
      "num_tokens": 34319342.0,
      "step": 19780
    },
    {
      "entropy": 5.757427787780761,
      "epoch": 1.5393503209492316,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047633494207809096,
      "loss": 5.2004,
      "mean_token_accuracy": 0.1859899252653122,
      "num_tokens": 34327468.0,
      "step": 19785
    },
    {
      "entropy": 5.656423711776734,
      "epoch": 1.539739350320949,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047632256767001977,
      "loss": 5.1305,
      "mean_token_accuracy": 0.19786737859249115,
      "num_tokens": 34336484.0,
      "step": 19790
    },
    {
      "entropy": 5.562729120254517,
      "epoch": 1.5401283796926668,
      "grad_norm": 1.140625,
      "learning_rate": 0.000476310190207199,
      "loss": 4.9878,
      "mean_token_accuracy": 0.20192447900772095,
      "num_tokens": 34344708.0,
      "step": 19795
    },
    {
      "entropy": 5.651788806915283,
      "epoch": 1.5405174090643845,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047629780968981653,
      "loss": 5.048,
      "mean_token_accuracy": 0.19703058451414107,
      "num_tokens": 34352757.0,
      "step": 19800
    },
    {
      "entropy": 5.604562044143677,
      "epoch": 1.540906438436102,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047628542611806007,
      "loss": 4.9994,
      "mean_token_accuracy": 0.203849096596241,
      "num_tokens": 34362070.0,
      "step": 19805
    },
    {
      "entropy": 5.6113992691040036,
      "epoch": 1.5412954678078195,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047627303949211773,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19022906422615052,
      "num_tokens": 34370136.0,
      "step": 19810
    },
    {
      "entropy": 5.557386159896851,
      "epoch": 1.5416844971795371,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004762606498121774,
      "loss": 5.0068,
      "mean_token_accuracy": 0.1962899461388588,
      "num_tokens": 34378534.0,
      "step": 19815
    },
    {
      "entropy": 5.60892539024353,
      "epoch": 1.5420735265512546,
      "grad_norm": 1.109375,
      "learning_rate": 0.000476248257078427,
      "loss": 5.0131,
      "mean_token_accuracy": 0.1980249762535095,
      "num_tokens": 34386313.0,
      "step": 19820
    },
    {
      "entropy": 5.614216947555542,
      "epoch": 1.542462555922972,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004762358612910547,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20082987248897552,
      "num_tokens": 34394566.0,
      "step": 19825
    },
    {
      "entropy": 5.537501096725464,
      "epoch": 1.5428515852946898,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004762234624502484,
      "loss": 4.954,
      "mean_token_accuracy": 0.20152128040790557,
      "num_tokens": 34403177.0,
      "step": 19830
    },
    {
      "entropy": 5.642577362060547,
      "epoch": 1.5432406146664073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047621106055619644,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19629845172166824,
      "num_tokens": 34411577.0,
      "step": 19835
    },
    {
      "entropy": 5.60940957069397,
      "epoch": 1.5436296440381247,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047619865560908687,
      "loss": 5.1029,
      "mean_token_accuracy": 0.18703222423791885,
      "num_tokens": 34420649.0,
      "step": 19840
    },
    {
      "entropy": 5.614952039718628,
      "epoch": 1.5440186734098424,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004761862476091079,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20465184450149537,
      "num_tokens": 34429778.0,
      "step": 19845
    },
    {
      "entropy": 5.649176073074341,
      "epoch": 1.5444077027815601,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047617383655644785,
      "loss": 5.0897,
      "mean_token_accuracy": 0.19585754573345185,
      "num_tokens": 34438412.0,
      "step": 19850
    },
    {
      "entropy": 5.652162981033325,
      "epoch": 1.5447967321532776,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004761614224512951,
      "loss": 5.0679,
      "mean_token_accuracy": 0.19063524454832076,
      "num_tokens": 34447022.0,
      "step": 19855
    },
    {
      "entropy": 5.639983606338501,
      "epoch": 1.545185761524995,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004761490052938379,
      "loss": 5.1248,
      "mean_token_accuracy": 0.19328873008489608,
      "num_tokens": 34455502.0,
      "step": 19860
    },
    {
      "entropy": 5.579786491394043,
      "epoch": 1.5455747908967128,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004761365850842646,
      "loss": 5.0676,
      "mean_token_accuracy": 0.1949102446436882,
      "num_tokens": 34463682.0,
      "step": 19865
    },
    {
      "entropy": 5.602315139770508,
      "epoch": 1.5459638202684303,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004761241618227639,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20273495018482207,
      "num_tokens": 34471678.0,
      "step": 19870
    },
    {
      "entropy": 5.597301626205445,
      "epoch": 1.5463528496401477,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047611173550952414,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19366590976715087,
      "num_tokens": 34479830.0,
      "step": 19875
    },
    {
      "entropy": 5.630930995941162,
      "epoch": 1.5467418790118654,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047609930614473387,
      "loss": 5.0893,
      "mean_token_accuracy": 0.19252397119998932,
      "num_tokens": 34487445.0,
      "step": 19880
    },
    {
      "entropy": 5.588702774047851,
      "epoch": 1.547130908383583,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047608687372858175,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20167562663555144,
      "num_tokens": 34496306.0,
      "step": 19885
    },
    {
      "entropy": 5.627710437774658,
      "epoch": 1.5475199377553004,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047607443826125633,
      "loss": 5.0755,
      "mean_token_accuracy": 0.19997381120920182,
      "num_tokens": 34505067.0,
      "step": 19890
    },
    {
      "entropy": 5.688541984558105,
      "epoch": 1.547908967127018,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047606199974294637,
      "loss": 5.1611,
      "mean_token_accuracy": 0.1870536118745804,
      "num_tokens": 34513947.0,
      "step": 19895
    },
    {
      "entropy": 5.7549717903137205,
      "epoch": 1.5482979964987358,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004760495581738406,
      "loss": 5.1624,
      "mean_token_accuracy": 0.18810292780399324,
      "num_tokens": 34523383.0,
      "step": 19900
    },
    {
      "entropy": 5.6299127578735355,
      "epoch": 1.5486870258704533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004760371135541278,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2018741860985756,
      "num_tokens": 34532376.0,
      "step": 19905
    },
    {
      "entropy": 5.684522485733032,
      "epoch": 1.5490760552421707,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004760246658839967,
      "loss": 5.1699,
      "mean_token_accuracy": 0.18633520752191543,
      "num_tokens": 34541075.0,
      "step": 19910
    },
    {
      "entropy": 5.587778377532959,
      "epoch": 1.5494650846138884,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004760122151636364,
      "loss": 4.963,
      "mean_token_accuracy": 0.20957433581352233,
      "num_tokens": 34550296.0,
      "step": 19915
    },
    {
      "entropy": 5.609297370910644,
      "epoch": 1.549854113985606,
      "grad_norm": 1.125,
      "learning_rate": 0.0004759997613932356,
      "loss": 5.1089,
      "mean_token_accuracy": 0.19689117670059203,
      "num_tokens": 34559185.0,
      "step": 19920
    },
    {
      "entropy": 5.692551326751709,
      "epoch": 1.5502431433573234,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047598730457298335,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19162441045045853,
      "num_tokens": 34568018.0,
      "step": 19925
    },
    {
      "entropy": 5.694375133514404,
      "epoch": 1.550632172729041,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047597484470306866,
      "loss": 5.0984,
      "mean_token_accuracy": 0.2034860819578171,
      "num_tokens": 34576835.0,
      "step": 19930
    },
    {
      "entropy": 5.6617310523986815,
      "epoch": 1.5510212021007586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004759623817836806,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1997387245297432,
      "num_tokens": 34585426.0,
      "step": 19935
    },
    {
      "entropy": 5.641145038604736,
      "epoch": 1.551410231472476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004759499158150082,
      "loss": 5.1112,
      "mean_token_accuracy": 0.1886995479464531,
      "num_tokens": 34594428.0,
      "step": 19940
    },
    {
      "entropy": 5.632831811904907,
      "epoch": 1.5517992608441937,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047593744679724076,
      "loss": 5.1186,
      "mean_token_accuracy": 0.19465104192495347,
      "num_tokens": 34603736.0,
      "step": 19945
    },
    {
      "entropy": 5.688205337524414,
      "epoch": 1.5521882902159114,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047592497473056733,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19554563164710997,
      "num_tokens": 34613295.0,
      "step": 19950
    },
    {
      "entropy": 5.619942045211792,
      "epoch": 1.552577319587629,
      "grad_norm": 1.125,
      "learning_rate": 0.00047591249961517724,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1928514465689659,
      "num_tokens": 34622955.0,
      "step": 19955
    },
    {
      "entropy": 5.584154558181763,
      "epoch": 1.5529663489593464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004759000214512598,
      "loss": 5.0618,
      "mean_token_accuracy": 0.19713790565729142,
      "num_tokens": 34631866.0,
      "step": 19960
    },
    {
      "entropy": 5.630996179580689,
      "epoch": 1.553355378331064,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004758875402390042,
      "loss": 5.1424,
      "mean_token_accuracy": 0.19057173132896424,
      "num_tokens": 34640281.0,
      "step": 19965
    },
    {
      "entropy": 5.664141035079956,
      "epoch": 1.5537444077027815,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047587505597859996,
      "loss": 5.0998,
      "mean_token_accuracy": 0.18910513818264008,
      "num_tokens": 34649080.0,
      "step": 19970
    },
    {
      "entropy": 5.615337276458741,
      "epoch": 1.554133437074499,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047586256867023646,
      "loss": 5.0594,
      "mean_token_accuracy": 0.19763769060373307,
      "num_tokens": 34657969.0,
      "step": 19975
    },
    {
      "entropy": 5.658559846878052,
      "epoch": 1.5545224664462167,
      "grad_norm": 1.125,
      "learning_rate": 0.0004758500783141032,
      "loss": 5.2011,
      "mean_token_accuracy": 0.1901307925581932,
      "num_tokens": 34667187.0,
      "step": 19980
    },
    {
      "entropy": 5.599294185638428,
      "epoch": 1.5549114958179344,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004758375849103897,
      "loss": 5.0445,
      "mean_token_accuracy": 0.19446507096290588,
      "num_tokens": 34676292.0,
      "step": 19985
    },
    {
      "entropy": 5.671174430847168,
      "epoch": 1.5553005251896517,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004758250884592855,
      "loss": 5.1537,
      "mean_token_accuracy": 0.19104479104280472,
      "num_tokens": 34685198.0,
      "step": 19990
    },
    {
      "entropy": 5.611014795303345,
      "epoch": 1.5556895545613694,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047581258896098024,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20463384538888932,
      "num_tokens": 34694019.0,
      "step": 19995
    },
    {
      "entropy": 5.653257226943969,
      "epoch": 1.556078583933087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004758000864156636,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1878348708152771,
      "num_tokens": 34703183.0,
      "step": 20000
    },
    {
      "entropy": 5.610570430755615,
      "epoch": 1.5564676133048045,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047578758082352527,
      "loss": 5.0651,
      "mean_token_accuracy": 0.19436854124069214,
      "num_tokens": 34711049.0,
      "step": 20005
    },
    {
      "entropy": 5.596108388900757,
      "epoch": 1.556856642676522,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047577507218475487,
      "loss": 5.1706,
      "mean_token_accuracy": 0.18169620335102082,
      "num_tokens": 34720130.0,
      "step": 20010
    },
    {
      "entropy": 5.667824411392212,
      "epoch": 1.5572456720482397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047576256049954236,
      "loss": 5.1065,
      "mean_token_accuracy": 0.1975086435675621,
      "num_tokens": 34728791.0,
      "step": 20015
    },
    {
      "entropy": 5.532134199142456,
      "epoch": 1.5576347014199572,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004757500457680776,
      "loss": 4.9013,
      "mean_token_accuracy": 0.20572284758090972,
      "num_tokens": 34737470.0,
      "step": 20020
    },
    {
      "entropy": 5.587539148330689,
      "epoch": 1.5580237307916747,
      "grad_norm": 1.125,
      "learning_rate": 0.0004757375279905504,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1956833854317665,
      "num_tokens": 34745774.0,
      "step": 20025
    },
    {
      "entropy": 5.55048017501831,
      "epoch": 1.5584127601633924,
      "grad_norm": 1.125,
      "learning_rate": 0.00047572500716715075,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20137026011943818,
      "num_tokens": 34753649.0,
      "step": 20030
    },
    {
      "entropy": 5.635546636581421,
      "epoch": 1.55880178953511,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047571248329806855,
      "loss": 5.0549,
      "mean_token_accuracy": 0.20032343864440919,
      "num_tokens": 34761833.0,
      "step": 20035
    },
    {
      "entropy": 5.560586309432983,
      "epoch": 1.5591908189068273,
      "grad_norm": 1.125,
      "learning_rate": 0.00047569995638349383,
      "loss": 5.0461,
      "mean_token_accuracy": 0.19266276955604553,
      "num_tokens": 34770969.0,
      "step": 20040
    },
    {
      "entropy": 5.615012264251709,
      "epoch": 1.559579848278545,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004756874264236168,
      "loss": 5.0949,
      "mean_token_accuracy": 0.19187907874584198,
      "num_tokens": 34779860.0,
      "step": 20045
    },
    {
      "entropy": 5.628459739685058,
      "epoch": 1.5599688776502627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047567489341862753,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20231844037771224,
      "num_tokens": 34788816.0,
      "step": 20050
    },
    {
      "entropy": 5.574292421340942,
      "epoch": 1.5603579070219802,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047566235736871607,
      "loss": 5.1048,
      "mean_token_accuracy": 0.19592037051916122,
      "num_tokens": 34797784.0,
      "step": 20055
    },
    {
      "entropy": 5.5566627979278564,
      "epoch": 1.5607469363936977,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047564981827407277,
      "loss": 4.9533,
      "mean_token_accuracy": 0.20971226543188096,
      "num_tokens": 34806553.0,
      "step": 20060
    },
    {
      "entropy": 5.604445600509644,
      "epoch": 1.5611359657654154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047563727613488785,
      "loss": 4.9779,
      "mean_token_accuracy": 0.19866746366024018,
      "num_tokens": 34814854.0,
      "step": 20065
    },
    {
      "entropy": 5.575188541412354,
      "epoch": 1.5615249951371328,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047562473095135154,
      "loss": 5.1659,
      "mean_token_accuracy": 0.19507531076669693,
      "num_tokens": 34823147.0,
      "step": 20070
    },
    {
      "entropy": 5.617303705215454,
      "epoch": 1.5619140245088503,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004756121827236544,
      "loss": 5.0542,
      "mean_token_accuracy": 0.20313065648078918,
      "num_tokens": 34831335.0,
      "step": 20075
    },
    {
      "entropy": 5.670486211776733,
      "epoch": 1.562303053880568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004755996314519866,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1911100387573242,
      "num_tokens": 34840118.0,
      "step": 20080
    },
    {
      "entropy": 5.620820426940918,
      "epoch": 1.5626920832522857,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004755870771365387,
      "loss": 5.0618,
      "mean_token_accuracy": 0.199837763607502,
      "num_tokens": 34848851.0,
      "step": 20085
    },
    {
      "entropy": 5.60090594291687,
      "epoch": 1.563081112624003,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047557451977750113,
      "loss": 4.9818,
      "mean_token_accuracy": 0.20136364549398422,
      "num_tokens": 34857660.0,
      "step": 20090
    },
    {
      "entropy": 5.588134098052978,
      "epoch": 1.5634701419957207,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004755619593750645,
      "loss": 5.0267,
      "mean_token_accuracy": 0.19643163532018662,
      "num_tokens": 34866828.0,
      "step": 20095
    },
    {
      "entropy": 5.64730281829834,
      "epoch": 1.5638591713674383,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004755493959294194,
      "loss": 5.1713,
      "mean_token_accuracy": 0.19020949751138688,
      "num_tokens": 34875461.0,
      "step": 20100
    },
    {
      "entropy": 5.545527839660645,
      "epoch": 1.5642482007391558,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004755368294407564,
      "loss": 5.0173,
      "mean_token_accuracy": 0.1919939249753952,
      "num_tokens": 34883914.0,
      "step": 20105
    },
    {
      "entropy": 5.632512044906616,
      "epoch": 1.5646372301108733,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004755242599092662,
      "loss": 5.1638,
      "mean_token_accuracy": 0.19406476616859436,
      "num_tokens": 34892191.0,
      "step": 20110
    },
    {
      "entropy": 5.657356262207031,
      "epoch": 1.565026259482591,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047551168733513956,
      "loss": 5.1122,
      "mean_token_accuracy": 0.19162083119153978,
      "num_tokens": 34900450.0,
      "step": 20115
    },
    {
      "entropy": 5.64314751625061,
      "epoch": 1.5654152888543085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047549911171856717,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2022545412182808,
      "num_tokens": 34908706.0,
      "step": 20120
    },
    {
      "entropy": 5.635372543334961,
      "epoch": 1.565804318226026,
      "grad_norm": 1.125,
      "learning_rate": 0.0004754865330597398,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19762219488620758,
      "num_tokens": 34916598.0,
      "step": 20125
    },
    {
      "entropy": 5.611127948760986,
      "epoch": 1.5661933475977436,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047547395135884854,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20053416341543198,
      "num_tokens": 34925850.0,
      "step": 20130
    },
    {
      "entropy": 5.658044672012329,
      "epoch": 1.5665823769694613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004754613666160841,
      "loss": 5.1303,
      "mean_token_accuracy": 0.19339323192834854,
      "num_tokens": 34934919.0,
      "step": 20135
    },
    {
      "entropy": 5.740956735610962,
      "epoch": 1.5669714063411786,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047544877883163753,
      "loss": 5.2327,
      "mean_token_accuracy": 0.18840365409851073,
      "num_tokens": 34943875.0,
      "step": 20140
    },
    {
      "entropy": 5.628398132324219,
      "epoch": 1.5673604357128963,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047543618800569975,
      "loss": 5.044,
      "mean_token_accuracy": 0.19426162242889405,
      "num_tokens": 34952458.0,
      "step": 20145
    },
    {
      "entropy": 5.595248460769653,
      "epoch": 1.567749465084614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047542359413846184,
      "loss": 5.1207,
      "mean_token_accuracy": 0.19421515017747878,
      "num_tokens": 34961550.0,
      "step": 20150
    },
    {
      "entropy": 5.6541748046875,
      "epoch": 1.5681384944563315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004754109972301149,
      "loss": 5.0433,
      "mean_token_accuracy": 0.1932863101363182,
      "num_tokens": 34969561.0,
      "step": 20155
    },
    {
      "entropy": 5.658156681060791,
      "epoch": 1.568527523828049,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047539839728085007,
      "loss": 5.1095,
      "mean_token_accuracy": 0.19819243252277374,
      "num_tokens": 34977904.0,
      "step": 20160
    },
    {
      "entropy": 5.614820718765259,
      "epoch": 1.5689165531997666,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004753857942908585,
      "loss": 5.137,
      "mean_token_accuracy": 0.19123504757881166,
      "num_tokens": 34986646.0,
      "step": 20165
    },
    {
      "entropy": 5.656961441040039,
      "epoch": 1.5693055825714841,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004753731882603315,
      "loss": 5.14,
      "mean_token_accuracy": 0.1924798533320427,
      "num_tokens": 34995387.0,
      "step": 20170
    },
    {
      "entropy": 5.663076162338257,
      "epoch": 1.5696946119432016,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047536057918946026,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20349431037902832,
      "num_tokens": 35004668.0,
      "step": 20175
    },
    {
      "entropy": 5.554481458663941,
      "epoch": 1.5700836413149193,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004753479670784361,
      "loss": 4.9396,
      "mean_token_accuracy": 0.20388782471418382,
      "num_tokens": 35012881.0,
      "step": 20180
    },
    {
      "entropy": 5.678840398788452,
      "epoch": 1.570472670686637,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004753353519274505,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1893852487206459,
      "num_tokens": 35022127.0,
      "step": 20185
    },
    {
      "entropy": 5.618812942504883,
      "epoch": 1.5708617000583545,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004753227337366948,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20334372371435167,
      "num_tokens": 35030066.0,
      "step": 20190
    },
    {
      "entropy": 5.5314621925354,
      "epoch": 1.571250729430072,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047531011250636046,
      "loss": 5.003,
      "mean_token_accuracy": 0.20537232160568236,
      "num_tokens": 35038560.0,
      "step": 20195
    },
    {
      "entropy": 5.604966735839843,
      "epoch": 1.5716397588017896,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047529748823663896,
      "loss": 5.1789,
      "mean_token_accuracy": 0.18840333372354506,
      "num_tokens": 35048011.0,
      "step": 20200
    },
    {
      "entropy": 5.685481452941895,
      "epoch": 1.572028788173507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004752848609277219,
      "loss": 5.1393,
      "mean_token_accuracy": 0.19389107078313828,
      "num_tokens": 35056247.0,
      "step": 20205
    },
    {
      "entropy": 5.650147151947022,
      "epoch": 1.5724178175452246,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047527223057980086,
      "loss": 5.0578,
      "mean_token_accuracy": 0.19735452085733413,
      "num_tokens": 35064885.0,
      "step": 20210
    },
    {
      "entropy": 5.613643074035645,
      "epoch": 1.5728068469169423,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047525959719306736,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1938127711415291,
      "num_tokens": 35073452.0,
      "step": 20215
    },
    {
      "entropy": 5.6129778861999515,
      "epoch": 1.5731958762886598,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004752469607677134,
      "loss": 5.1524,
      "mean_token_accuracy": 0.19291624426841736,
      "num_tokens": 35082459.0,
      "step": 20220
    },
    {
      "entropy": 5.6289222240448,
      "epoch": 1.5735849056603772,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004752343213039305,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19284766912460327,
      "num_tokens": 35091336.0,
      "step": 20225
    },
    {
      "entropy": 5.611125373840332,
      "epoch": 1.573973935032095,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004752216788019104,
      "loss": 5.0556,
      "mean_token_accuracy": 0.20227845013141632,
      "num_tokens": 35099802.0,
      "step": 20230
    },
    {
      "entropy": 5.633749055862427,
      "epoch": 1.5743629644038126,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004752090332618451,
      "loss": 5.0898,
      "mean_token_accuracy": 0.19005305022001268,
      "num_tokens": 35108520.0,
      "step": 20235
    },
    {
      "entropy": 5.644745302200318,
      "epoch": 1.57475199377553,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004751963846839263,
      "loss": 5.1227,
      "mean_token_accuracy": 0.19701891988515854,
      "num_tokens": 35117481.0,
      "step": 20240
    },
    {
      "entropy": 5.585175514221191,
      "epoch": 1.5751410231472476,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047518373306834605,
      "loss": 5.0798,
      "mean_token_accuracy": 0.1909394234418869,
      "num_tokens": 35126605.0,
      "step": 20245
    },
    {
      "entropy": 5.584819746017456,
      "epoch": 1.5755300525189653,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047517107841529626,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20656824111938477,
      "num_tokens": 35135312.0,
      "step": 20250
    },
    {
      "entropy": 5.626561498641967,
      "epoch": 1.5759190818906827,
      "grad_norm": 1.0625,
      "learning_rate": 0.000475158420724969,
      "loss": 5.0434,
      "mean_token_accuracy": 0.19772955179214477,
      "num_tokens": 35144759.0,
      "step": 20255
    },
    {
      "entropy": 5.642653942108154,
      "epoch": 1.5763081112624002,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047514575999755627,
      "loss": 5.1632,
      "mean_token_accuracy": 0.19583328068256378,
      "num_tokens": 35153701.0,
      "step": 20260
    },
    {
      "entropy": 5.661393547058106,
      "epoch": 1.576697140634118,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047513309623325024,
      "loss": 5.204,
      "mean_token_accuracy": 0.18794293850660324,
      "num_tokens": 35162503.0,
      "step": 20265
    },
    {
      "entropy": 5.596063184738159,
      "epoch": 1.5770861700058354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004751204294322429,
      "loss": 5.05,
      "mean_token_accuracy": 0.19753761887550353,
      "num_tokens": 35171209.0,
      "step": 20270
    },
    {
      "entropy": 5.591230726242065,
      "epoch": 1.5774751993775529,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047510775959472675,
      "loss": 5.0676,
      "mean_token_accuracy": 0.19537246525287627,
      "num_tokens": 35179623.0,
      "step": 20275
    },
    {
      "entropy": 5.600982236862182,
      "epoch": 1.5778642287492706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004750950867208937,
      "loss": 5.0554,
      "mean_token_accuracy": 0.19791653007268906,
      "num_tokens": 35188679.0,
      "step": 20280
    },
    {
      "entropy": 5.587707567214966,
      "epoch": 1.5782532581209883,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004750824108109362,
      "loss": 5.0142,
      "mean_token_accuracy": 0.1996718868613243,
      "num_tokens": 35197325.0,
      "step": 20285
    },
    {
      "entropy": 5.677344608306885,
      "epoch": 1.5786422874927057,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004750697318650466,
      "loss": 5.1607,
      "mean_token_accuracy": 0.18829642236232758,
      "num_tokens": 35206741.0,
      "step": 20290
    },
    {
      "entropy": 5.637091588973999,
      "epoch": 1.5790313168644232,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004750570498834173,
      "loss": 5.095,
      "mean_token_accuracy": 0.19263596832752228,
      "num_tokens": 35215346.0,
      "step": 20295
    },
    {
      "entropy": 5.633522939682007,
      "epoch": 1.579420346236141,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004750443648662407,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19598385840654373,
      "num_tokens": 35223811.0,
      "step": 20300
    },
    {
      "entropy": 5.646094274520874,
      "epoch": 1.5798093756078584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047503167681370924,
      "loss": 5.1121,
      "mean_token_accuracy": 0.18771164417266845,
      "num_tokens": 35232438.0,
      "step": 20305
    },
    {
      "entropy": 5.646022653579712,
      "epoch": 1.5801984049795759,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004750189857260154,
      "loss": 5.0735,
      "mean_token_accuracy": 0.1912990003824234,
      "num_tokens": 35242529.0,
      "step": 20310
    },
    {
      "entropy": 5.535845232009888,
      "epoch": 1.5805874343512936,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004750062916033519,
      "loss": 4.9248,
      "mean_token_accuracy": 0.20278697311878205,
      "num_tokens": 35250695.0,
      "step": 20315
    },
    {
      "entropy": 5.626095819473266,
      "epoch": 1.580976463723011,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004749935944459113,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20065564811229705,
      "num_tokens": 35259212.0,
      "step": 20320
    },
    {
      "entropy": 5.573158311843872,
      "epoch": 1.5813654930947285,
      "grad_norm": 1.125,
      "learning_rate": 0.0004749808942538862,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20158760994672775,
      "num_tokens": 35267667.0,
      "step": 20325
    },
    {
      "entropy": 5.627477836608887,
      "epoch": 1.5817545224664462,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004749681910274693,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19469761699438096,
      "num_tokens": 35275850.0,
      "step": 20330
    },
    {
      "entropy": 5.688288640975952,
      "epoch": 1.582143551838164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047495548476685334,
      "loss": 5.0562,
      "mean_token_accuracy": 0.194144044816494,
      "num_tokens": 35284635.0,
      "step": 20335
    },
    {
      "entropy": 5.677833461761475,
      "epoch": 1.5825325812098814,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047494277547223125,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20119666308164597,
      "num_tokens": 35292335.0,
      "step": 20340
    },
    {
      "entropy": 5.58170485496521,
      "epoch": 1.5829216105815989,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047493006314379573,
      "loss": 5.0691,
      "mean_token_accuracy": 0.19483873397111892,
      "num_tokens": 35300709.0,
      "step": 20345
    },
    {
      "entropy": 5.632246732711792,
      "epoch": 1.5833106399533166,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047491734778173976,
      "loss": 5.046,
      "mean_token_accuracy": 0.2037111535668373,
      "num_tokens": 35308956.0,
      "step": 20350
    },
    {
      "entropy": 5.5469707489013675,
      "epoch": 1.583699669325034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047490462938625617,
      "loss": 5.0718,
      "mean_token_accuracy": 0.1930500864982605,
      "num_tokens": 35317663.0,
      "step": 20355
    },
    {
      "entropy": 5.646196269989014,
      "epoch": 1.5840886986967515,
      "grad_norm": 1.0,
      "learning_rate": 0.00047489190795753806,
      "loss": 5.1422,
      "mean_token_accuracy": 0.1858256995677948,
      "num_tokens": 35327412.0,
      "step": 20360
    },
    {
      "entropy": 5.715460348129272,
      "epoch": 1.5844777280684692,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004748791834957784,
      "loss": 5.1611,
      "mean_token_accuracy": 0.18700429797172546,
      "num_tokens": 35335888.0,
      "step": 20365
    },
    {
      "entropy": 5.649503755569458,
      "epoch": 1.5848667574401867,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047486645600117037,
      "loss": 5.0933,
      "mean_token_accuracy": 0.20034096986055375,
      "num_tokens": 35345122.0,
      "step": 20370
    },
    {
      "entropy": 5.621166610717774,
      "epoch": 1.5852557868119042,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004748537254739068,
      "loss": 5.1348,
      "mean_token_accuracy": 0.1947360649704933,
      "num_tokens": 35353751.0,
      "step": 20375
    },
    {
      "entropy": 5.638257169723511,
      "epoch": 1.5856448161836219,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004748409919141812,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20239398777484893,
      "num_tokens": 35363083.0,
      "step": 20380
    },
    {
      "entropy": 5.590206956863403,
      "epoch": 1.5860338455553395,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004748282553221865,
      "loss": 5.0339,
      "mean_token_accuracy": 0.2062219649553299,
      "num_tokens": 35371198.0,
      "step": 20385
    },
    {
      "entropy": 5.6710021018981935,
      "epoch": 1.586422874927057,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004748155156981162,
      "loss": 5.1205,
      "mean_token_accuracy": 0.19795178920030593,
      "num_tokens": 35379633.0,
      "step": 20390
    },
    {
      "entropy": 5.627018594741822,
      "epoch": 1.5868119042987745,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047480277304216346,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19585523903369903,
      "num_tokens": 35388172.0,
      "step": 20395
    },
    {
      "entropy": 5.587087678909302,
      "epoch": 1.5872009336704922,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047479002735452164,
      "loss": 5.0545,
      "mean_token_accuracy": 0.19371065348386765,
      "num_tokens": 35397126.0,
      "step": 20400
    },
    {
      "entropy": 5.615014600753784,
      "epoch": 1.5875899630422097,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047477727863538415,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2033095419406891,
      "num_tokens": 35405364.0,
      "step": 20405
    },
    {
      "entropy": 5.561865997314453,
      "epoch": 1.5879789924139271,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047476452688494444,
      "loss": 5.0312,
      "mean_token_accuracy": 0.1998643234372139,
      "num_tokens": 35414187.0,
      "step": 20410
    },
    {
      "entropy": 5.612519550323486,
      "epoch": 1.5883680217856448,
      "grad_norm": 1.171875,
      "learning_rate": 0.000474751772103396,
      "loss": 5.0258,
      "mean_token_accuracy": 0.197518290579319,
      "num_tokens": 35422694.0,
      "step": 20415
    },
    {
      "entropy": 5.655950927734375,
      "epoch": 1.5887570511573625,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004747390142909323,
      "loss": 5.1678,
      "mean_token_accuracy": 0.1940804421901703,
      "num_tokens": 35431176.0,
      "step": 20420
    },
    {
      "entropy": 5.63248119354248,
      "epoch": 1.5891460805290798,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047472625344774713,
      "loss": 5.096,
      "mean_token_accuracy": 0.18586088418960572,
      "num_tokens": 35439938.0,
      "step": 20425
    },
    {
      "entropy": 5.650304889678955,
      "epoch": 1.5895351099007975,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047471348957403374,
      "loss": 5.0455,
      "mean_token_accuracy": 0.19342994540929795,
      "num_tokens": 35447864.0,
      "step": 20430
    },
    {
      "entropy": 5.67044587135315,
      "epoch": 1.5899241392725152,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004747007226699862,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19584671258926392,
      "num_tokens": 35456047.0,
      "step": 20435
    },
    {
      "entropy": 5.568166255950928,
      "epoch": 1.5903131686442327,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047468795273579803,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20492106825113296,
      "num_tokens": 35464510.0,
      "step": 20440
    },
    {
      "entropy": 5.609268093109131,
      "epoch": 1.5907021980159501,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004746751797716629,
      "loss": 5.12,
      "mean_token_accuracy": 0.18894250988960265,
      "num_tokens": 35472715.0,
      "step": 20445
    },
    {
      "entropy": 5.598772192001343,
      "epoch": 1.5910912273876678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004746624037777748,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20071628987789153,
      "num_tokens": 35482094.0,
      "step": 20450
    },
    {
      "entropy": 5.627783632278442,
      "epoch": 1.5914802567593853,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047464962475432754,
      "loss": 5.0651,
      "mean_token_accuracy": 0.1978485882282257,
      "num_tokens": 35489864.0,
      "step": 20455
    },
    {
      "entropy": 5.660768747329712,
      "epoch": 1.5918692861311028,
      "grad_norm": 1.171875,
      "learning_rate": 0.000474636842701515,
      "loss": 5.1286,
      "mean_token_accuracy": 0.1918771132826805,
      "num_tokens": 35498840.0,
      "step": 20460
    },
    {
      "entropy": 5.543379306793213,
      "epoch": 1.5922583155028205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004746240576195311,
      "loss": 4.9764,
      "mean_token_accuracy": 0.19892841577529907,
      "num_tokens": 35506954.0,
      "step": 20465
    },
    {
      "entropy": 5.603975582122803,
      "epoch": 1.5926473448745382,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047461126950856987,
      "loss": 5.0927,
      "mean_token_accuracy": 0.19669529497623445,
      "num_tokens": 35515796.0,
      "step": 20470
    },
    {
      "entropy": 5.63947639465332,
      "epoch": 1.5930363742462554,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004745984783688253,
      "loss": 5.043,
      "mean_token_accuracy": 0.20249154418706894,
      "num_tokens": 35523713.0,
      "step": 20475
    },
    {
      "entropy": 5.66815676689148,
      "epoch": 1.5934254036179731,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047458568420049153,
      "loss": 5.1822,
      "mean_token_accuracy": 0.18614078015089036,
      "num_tokens": 35532800.0,
      "step": 20480
    },
    {
      "entropy": 5.667867040634155,
      "epoch": 1.5938144329896908,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047457288700376274,
      "loss": 5.0793,
      "mean_token_accuracy": 0.20315959453582763,
      "num_tokens": 35540865.0,
      "step": 20485
    },
    {
      "entropy": 5.668543148040771,
      "epoch": 1.5942034623614083,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047456008677883304,
      "loss": 5.1658,
      "mean_token_accuracy": 0.19243755787611008,
      "num_tokens": 35549498.0,
      "step": 20490
    },
    {
      "entropy": 5.640258836746216,
      "epoch": 1.5945924917331258,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004745472835258966,
      "loss": 5.0791,
      "mean_token_accuracy": 0.1981320172548294,
      "num_tokens": 35557756.0,
      "step": 20495
    },
    {
      "entropy": 5.700021743774414,
      "epoch": 1.5949815211048435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047453447724514773,
      "loss": 5.153,
      "mean_token_accuracy": 0.19131090641021728,
      "num_tokens": 35566062.0,
      "step": 20500
    },
    {
      "entropy": 5.613107204437256,
      "epoch": 1.595370550476561,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745216679367808,
      "loss": 4.9752,
      "mean_token_accuracy": 0.198999260365963,
      "num_tokens": 35574377.0,
      "step": 20505
    },
    {
      "entropy": 5.589367628097534,
      "epoch": 1.5957595798482784,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004745088556009901,
      "loss": 5.0712,
      "mean_token_accuracy": 0.20105661153793336,
      "num_tokens": 35582254.0,
      "step": 20510
    },
    {
      "entropy": 5.625000762939453,
      "epoch": 1.5961486092199961,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004744960402379701,
      "loss": 5.1282,
      "mean_token_accuracy": 0.1921658918261528,
      "num_tokens": 35590804.0,
      "step": 20515
    },
    {
      "entropy": 5.641349267959595,
      "epoch": 1.5965376385917138,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047448322184791525,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20003836154937743,
      "num_tokens": 35599547.0,
      "step": 20520
    },
    {
      "entropy": 5.73191089630127,
      "epoch": 1.596926667963431,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047447040043101994,
      "loss": 5.1648,
      "mean_token_accuracy": 0.19129930287599564,
      "num_tokens": 35608948.0,
      "step": 20525
    },
    {
      "entropy": 5.732142543792724,
      "epoch": 1.5973156973351488,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047445757598747886,
      "loss": 5.2445,
      "mean_token_accuracy": 0.18365987092256547,
      "num_tokens": 35618243.0,
      "step": 20530
    },
    {
      "entropy": 5.611090230941772,
      "epoch": 1.5977047267068665,
      "grad_norm": 1.125,
      "learning_rate": 0.0004744447485174864,
      "loss": 5.0417,
      "mean_token_accuracy": 0.2022300660610199,
      "num_tokens": 35626938.0,
      "step": 20535
    },
    {
      "entropy": 5.656848669052124,
      "epoch": 1.598093756078584,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047443191802123746,
      "loss": 5.1043,
      "mean_token_accuracy": 0.20227303802967073,
      "num_tokens": 35635612.0,
      "step": 20540
    },
    {
      "entropy": 5.628291320800781,
      "epoch": 1.5984827854503014,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047441908449892664,
      "loss": 5.0893,
      "mean_token_accuracy": 0.19863858819007874,
      "num_tokens": 35644320.0,
      "step": 20545
    },
    {
      "entropy": 5.595477676391601,
      "epoch": 1.5988718148220191,
      "grad_norm": 1.125,
      "learning_rate": 0.00047440624795074855,
      "loss": 5.0352,
      "mean_token_accuracy": 0.19717060029506683,
      "num_tokens": 35652606.0,
      "step": 20550
    },
    {
      "entropy": 5.623254776000977,
      "epoch": 1.5992608441937366,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047439340837689804,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2002899944782257,
      "num_tokens": 35661082.0,
      "step": 20555
    },
    {
      "entropy": 5.602928733825683,
      "epoch": 1.599649873565454,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047438056577756986,
      "loss": 5.1169,
      "mean_token_accuracy": 0.1872546926140785,
      "num_tokens": 35669926.0,
      "step": 20560
    },
    {
      "entropy": 5.645620250701905,
      "epoch": 1.6000389029371718,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047436772015295903,
      "loss": 5.0949,
      "mean_token_accuracy": 0.2000372067093849,
      "num_tokens": 35678964.0,
      "step": 20565
    },
    {
      "entropy": 5.664347362518311,
      "epoch": 1.6004279323088895,
      "grad_norm": 1.125,
      "learning_rate": 0.00047435487150326036,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20405319780111314,
      "num_tokens": 35687943.0,
      "step": 20570
    },
    {
      "entropy": 5.598870277404785,
      "epoch": 1.6008169616806067,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004743420198286688,
      "loss": 5.0795,
      "mean_token_accuracy": 0.1985732063651085,
      "num_tokens": 35696684.0,
      "step": 20575
    },
    {
      "entropy": 5.583149433135986,
      "epoch": 1.6012059910523244,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047432916512937936,
      "loss": 5.1041,
      "mean_token_accuracy": 0.19321274608373643,
      "num_tokens": 35705038.0,
      "step": 20580
    },
    {
      "entropy": 5.699066495895385,
      "epoch": 1.6015950204240421,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004743163074055871,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19549170583486558,
      "num_tokens": 35713638.0,
      "step": 20585
    },
    {
      "entropy": 5.659934949874878,
      "epoch": 1.6019840497957596,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004743034466574871,
      "loss": 5.0844,
      "mean_token_accuracy": 0.19501836746931075,
      "num_tokens": 35721935.0,
      "step": 20590
    },
    {
      "entropy": 5.667053890228272,
      "epoch": 1.602373079167477,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004742905828852746,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19410964250564575,
      "num_tokens": 35731610.0,
      "step": 20595
    },
    {
      "entropy": 5.650221633911133,
      "epoch": 1.6027621085391948,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004742777160891447,
      "loss": 5.1214,
      "mean_token_accuracy": 0.19749108850955963,
      "num_tokens": 35740956.0,
      "step": 20600
    },
    {
      "entropy": 5.590960597991943,
      "epoch": 1.6031511379109122,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004742648462692926,
      "loss": 4.9904,
      "mean_token_accuracy": 0.203285613656044,
      "num_tokens": 35749316.0,
      "step": 20605
    },
    {
      "entropy": 5.650044536590576,
      "epoch": 1.6035401672826297,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047425197342591363,
      "loss": 5.172,
      "mean_token_accuracy": 0.19296573996543884,
      "num_tokens": 35758875.0,
      "step": 20610
    },
    {
      "entropy": 5.62503752708435,
      "epoch": 1.6039291966543474,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004742390975592031,
      "loss": 5.1026,
      "mean_token_accuracy": 0.20000848323106765,
      "num_tokens": 35767851.0,
      "step": 20615
    },
    {
      "entropy": 5.6983216285705565,
      "epoch": 1.604318226026065,
      "grad_norm": 1.125,
      "learning_rate": 0.00047422621866935645,
      "loss": 5.0135,
      "mean_token_accuracy": 0.19999344646930695,
      "num_tokens": 35776393.0,
      "step": 20620
    },
    {
      "entropy": 5.603093671798706,
      "epoch": 1.6047072553977826,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000474213336756569,
      "loss": 5.013,
      "mean_token_accuracy": 0.2031023919582367,
      "num_tokens": 35784868.0,
      "step": 20625
    },
    {
      "entropy": 5.596323728561401,
      "epoch": 1.6050962847695,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004742004518210362,
      "loss": 5.0955,
      "mean_token_accuracy": 0.18370085060596467,
      "num_tokens": 35793812.0,
      "step": 20630
    },
    {
      "entropy": 5.632270669937133,
      "epoch": 1.6054853141412178,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004741875638629536,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19073866605758666,
      "num_tokens": 35802206.0,
      "step": 20635
    },
    {
      "entropy": 5.619304943084717,
      "epoch": 1.6058743435129352,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004741746728825168,
      "loss": 5.0303,
      "mean_token_accuracy": 0.1967957228422165,
      "num_tokens": 35811095.0,
      "step": 20640
    },
    {
      "entropy": 5.726135158538819,
      "epoch": 1.6062633728846527,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004741617788799213,
      "loss": 5.1537,
      "mean_token_accuracy": 0.19018188267946243,
      "num_tokens": 35819799.0,
      "step": 20645
    },
    {
      "entropy": 5.573319673538208,
      "epoch": 1.6066524022563704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047414888185536285,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1965108558535576,
      "num_tokens": 35829130.0,
      "step": 20650
    },
    {
      "entropy": 5.578005456924439,
      "epoch": 1.6070414316280879,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004741359818090371,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19931916892528534,
      "num_tokens": 35838020.0,
      "step": 20655
    },
    {
      "entropy": 5.610688066482544,
      "epoch": 1.6074304609998054,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047412307874113976,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20839978456497193,
      "num_tokens": 35845843.0,
      "step": 20660
    },
    {
      "entropy": 5.557564401626587,
      "epoch": 1.607819490371523,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004741101726518667,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2021481290459633,
      "num_tokens": 35854505.0,
      "step": 20665
    },
    {
      "entropy": 5.611812496185303,
      "epoch": 1.6082085197432407,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004740972635414136,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19782837331295014,
      "num_tokens": 35862847.0,
      "step": 20670
    },
    {
      "entropy": 5.64195761680603,
      "epoch": 1.6085975491149582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004740843514099765,
      "loss": 5.0325,
      "mean_token_accuracy": 0.19501852095127106,
      "num_tokens": 35871471.0,
      "step": 20675
    },
    {
      "entropy": 5.630333757400512,
      "epoch": 1.6089865784866757,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004740714362577512,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19419005513191223,
      "num_tokens": 35880266.0,
      "step": 20680
    },
    {
      "entropy": 5.584820032119751,
      "epoch": 1.6093756078583934,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047405851808493364,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20448884814977647,
      "num_tokens": 35888727.0,
      "step": 20685
    },
    {
      "entropy": 5.579598379135132,
      "epoch": 1.6097646372301109,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047404559689172006,
      "loss": 5.0663,
      "mean_token_accuracy": 0.19719063937664033,
      "num_tokens": 35897750.0,
      "step": 20690
    },
    {
      "entropy": 5.587022209167481,
      "epoch": 1.6101536666018283,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047403267267830617,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20050567984580994,
      "num_tokens": 35906254.0,
      "step": 20695
    },
    {
      "entropy": 5.608237028121948,
      "epoch": 1.610542695973546,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047401974544488844,
      "loss": 5.1245,
      "mean_token_accuracy": 0.18764768093824385,
      "num_tokens": 35915180.0,
      "step": 20700
    },
    {
      "entropy": 5.682959747314453,
      "epoch": 1.6109317253452635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004740068151916628,
      "loss": 5.0786,
      "mean_token_accuracy": 0.1847858563065529,
      "num_tokens": 35924336.0,
      "step": 20705
    },
    {
      "entropy": 5.640835857391357,
      "epoch": 1.611320754716981,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004739938819188255,
      "loss": 5.1011,
      "mean_token_accuracy": 0.19750546663999557,
      "num_tokens": 35933031.0,
      "step": 20710
    },
    {
      "entropy": 5.600457954406738,
      "epoch": 1.6117097840886987,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004739809456265727,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20755631923675538,
      "num_tokens": 35941794.0,
      "step": 20715
    },
    {
      "entropy": 5.715494155883789,
      "epoch": 1.6120988134604164,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004739680063151008,
      "loss": 5.1784,
      "mean_token_accuracy": 0.190445214509964,
      "num_tokens": 35949636.0,
      "step": 20720
    },
    {
      "entropy": 5.6185595989227295,
      "epoch": 1.6124878428321339,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004739550639846061,
      "loss": 5.1768,
      "mean_token_accuracy": 0.19313769191503524,
      "num_tokens": 35958842.0,
      "step": 20725
    },
    {
      "entropy": 5.609982872009278,
      "epoch": 1.6128768722038513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047394211863528496,
      "loss": 5.0159,
      "mean_token_accuracy": 0.193978451192379,
      "num_tokens": 35967897.0,
      "step": 20730
    },
    {
      "entropy": 5.573044919967652,
      "epoch": 1.613265901575569,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004739291702673338,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2068732649087906,
      "num_tokens": 35976309.0,
      "step": 20735
    },
    {
      "entropy": 5.7254173278808596,
      "epoch": 1.6136549309472865,
      "grad_norm": 1.125,
      "learning_rate": 0.0004739162188809492,
      "loss": 5.1747,
      "mean_token_accuracy": 0.1895492836833,
      "num_tokens": 35984401.0,
      "step": 20740
    },
    {
      "entropy": 5.608604526519775,
      "epoch": 1.614043960319004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047390326447632747,
      "loss": 5.0583,
      "mean_token_accuracy": 0.191900572180748,
      "num_tokens": 35993068.0,
      "step": 20745
    },
    {
      "entropy": 5.720227432250977,
      "epoch": 1.6144329896907217,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004738903070536654,
      "loss": 5.21,
      "mean_token_accuracy": 0.19097689688205718,
      "num_tokens": 36001559.0,
      "step": 20750
    },
    {
      "entropy": 5.679964447021485,
      "epoch": 1.6148220190624392,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004738773466131594,
      "loss": 5.145,
      "mean_token_accuracy": 0.19234924912452697,
      "num_tokens": 36010675.0,
      "step": 20755
    },
    {
      "entropy": 5.54162449836731,
      "epoch": 1.6152110484341566,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004738643831550063,
      "loss": 4.8744,
      "mean_token_accuracy": 0.21470435261726378,
      "num_tokens": 36019498.0,
      "step": 20760
    },
    {
      "entropy": 5.522093009948731,
      "epoch": 1.6156000778058743,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004738514166794026,
      "loss": 4.958,
      "mean_token_accuracy": 0.21044186055660247,
      "num_tokens": 36028206.0,
      "step": 20765
    },
    {
      "entropy": 5.685051202774048,
      "epoch": 1.615989107177592,
      "grad_norm": 1.125,
      "learning_rate": 0.00047383844718654525,
      "loss": 5.15,
      "mean_token_accuracy": 0.199816232919693,
      "num_tokens": 36037341.0,
      "step": 20770
    },
    {
      "entropy": 5.622928237915039,
      "epoch": 1.6163781365493095,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004738254746766309,
      "loss": 5.0416,
      "mean_token_accuracy": 0.20137808322906495,
      "num_tokens": 36046596.0,
      "step": 20775
    },
    {
      "entropy": 5.617893886566162,
      "epoch": 1.616767165921027,
      "grad_norm": 1.125,
      "learning_rate": 0.0004738124991498565,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20008259564638137,
      "num_tokens": 36055031.0,
      "step": 20780
    },
    {
      "entropy": 5.688340091705323,
      "epoch": 1.6171561952927447,
      "grad_norm": 1.125,
      "learning_rate": 0.00047379952060641866,
      "loss": 5.141,
      "mean_token_accuracy": 0.19778156876564026,
      "num_tokens": 36064049.0,
      "step": 20785
    },
    {
      "entropy": 5.680044412612915,
      "epoch": 1.6175452246644622,
      "grad_norm": 1.125,
      "learning_rate": 0.00047378653904651475,
      "loss": 5.1364,
      "mean_token_accuracy": 0.18954210877418518,
      "num_tokens": 36072886.0,
      "step": 20790
    },
    {
      "entropy": 5.668096590042114,
      "epoch": 1.6179342540361796,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047377355447034136,
      "loss": 5.1337,
      "mean_token_accuracy": 0.19390322864055634,
      "num_tokens": 36082133.0,
      "step": 20795
    },
    {
      "entropy": 5.653011608123779,
      "epoch": 1.6183232834078973,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004737605668780958,
      "loss": 5.0456,
      "mean_token_accuracy": 0.19921970665454863,
      "num_tokens": 36090980.0,
      "step": 20800
    },
    {
      "entropy": 5.6601104736328125,
      "epoch": 1.618712312779615,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047374757626997494,
      "loss": 5.1008,
      "mean_token_accuracy": 0.19587356150150298,
      "num_tokens": 36099584.0,
      "step": 20805
    },
    {
      "entropy": 5.552107858657837,
      "epoch": 1.6191013421513323,
      "grad_norm": 1.25,
      "learning_rate": 0.0004737345826461759,
      "loss": 5.0497,
      "mean_token_accuracy": 0.19332421720027923,
      "num_tokens": 36108052.0,
      "step": 20810
    },
    {
      "entropy": 5.558855962753296,
      "epoch": 1.61949037152305,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004737215860068959,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20958798974752427,
      "num_tokens": 36116623.0,
      "step": 20815
    },
    {
      "entropy": 5.7140600204467775,
      "epoch": 1.6198794008947677,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047370858635233223,
      "loss": 5.0755,
      "mean_token_accuracy": 0.19617012441158294,
      "num_tokens": 36124953.0,
      "step": 20820
    },
    {
      "entropy": 5.632710218429565,
      "epoch": 1.6202684302664851,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047369558368268194,
      "loss": 5.0772,
      "mean_token_accuracy": 0.20054130107164383,
      "num_tokens": 36133500.0,
      "step": 20825
    },
    {
      "entropy": 5.59528374671936,
      "epoch": 1.6206574596382026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004736825779981424,
      "loss": 5.0165,
      "mean_token_accuracy": 0.199115851521492,
      "num_tokens": 36141934.0,
      "step": 20830
    },
    {
      "entropy": 5.584219455718994,
      "epoch": 1.6210464890099203,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004736695692989111,
      "loss": 5.0335,
      "mean_token_accuracy": 0.19806117415428162,
      "num_tokens": 36150592.0,
      "step": 20835
    },
    {
      "entropy": 5.609458160400391,
      "epoch": 1.6214355183816378,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004736565575851852,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20246375799179078,
      "num_tokens": 36159377.0,
      "step": 20840
    },
    {
      "entropy": 5.633948087692261,
      "epoch": 1.6218245477533553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004736435428571622,
      "loss": 5.0205,
      "mean_token_accuracy": 0.1959124207496643,
      "num_tokens": 36168159.0,
      "step": 20845
    },
    {
      "entropy": 5.553215169906617,
      "epoch": 1.622213577125073,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004736305251150397,
      "loss": 4.9989,
      "mean_token_accuracy": 0.19571877717971803,
      "num_tokens": 36176108.0,
      "step": 20850
    },
    {
      "entropy": 5.639344787597656,
      "epoch": 1.6226026064967907,
      "grad_norm": 1.25,
      "learning_rate": 0.0004736175043590151,
      "loss": 5.156,
      "mean_token_accuracy": 0.18885980248451234,
      "num_tokens": 36184930.0,
      "step": 20855
    },
    {
      "entropy": 5.695063686370849,
      "epoch": 1.622991635868508,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000473604480589286,
      "loss": 5.0821,
      "mean_token_accuracy": 0.196304526925087,
      "num_tokens": 36193308.0,
      "step": 20860
    },
    {
      "entropy": 5.618703842163086,
      "epoch": 1.6233806652402256,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047359145380605007,
      "loss": 5.0407,
      "mean_token_accuracy": 0.19834544509649277,
      "num_tokens": 36202016.0,
      "step": 20865
    },
    {
      "entropy": 5.59560341835022,
      "epoch": 1.6237696946119433,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004735784240095049,
      "loss": 5.0259,
      "mean_token_accuracy": 0.1984819307923317,
      "num_tokens": 36210757.0,
      "step": 20870
    },
    {
      "entropy": 5.542816686630249,
      "epoch": 1.6241587239836608,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047356539119984813,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20314572453498841,
      "num_tokens": 36219052.0,
      "step": 20875
    },
    {
      "entropy": 5.564661931991577,
      "epoch": 1.6245477533553783,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004735523553772777,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21042634546756744,
      "num_tokens": 36227890.0,
      "step": 20880
    },
    {
      "entropy": 5.623157596588134,
      "epoch": 1.624936782727096,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004735393165419912,
      "loss": 5.087,
      "mean_token_accuracy": 0.19788916558027267,
      "num_tokens": 36236725.0,
      "step": 20885
    },
    {
      "entropy": 5.693569850921631,
      "epoch": 1.6253258120988134,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047352627469418666,
      "loss": 5.0878,
      "mean_token_accuracy": 0.19651439487934114,
      "num_tokens": 36246220.0,
      "step": 20890
    },
    {
      "entropy": 5.701940536499023,
      "epoch": 1.625714841470531,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004735132298340619,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19450931400060653,
      "num_tokens": 36254662.0,
      "step": 20895
    },
    {
      "entropy": 5.608836936950683,
      "epoch": 1.6261038708422486,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004735001819618148,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19978391975164414,
      "num_tokens": 36262446.0,
      "step": 20900
    },
    {
      "entropy": 5.678788661956787,
      "epoch": 1.6264929002139663,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004734871310776434,
      "loss": 5.098,
      "mean_token_accuracy": 0.19005067199468612,
      "num_tokens": 36271284.0,
      "step": 20905
    },
    {
      "entropy": 5.586025524139404,
      "epoch": 1.6268819295856836,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004734740771817457,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20308156311511993,
      "num_tokens": 36279180.0,
      "step": 20910
    },
    {
      "entropy": 5.585681962966919,
      "epoch": 1.6272709589574013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004734610202743198,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20034157633781433,
      "num_tokens": 36287714.0,
      "step": 20915
    },
    {
      "entropy": 5.668402290344238,
      "epoch": 1.627659988329119,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004734479603555638,
      "loss": 5.0939,
      "mean_token_accuracy": 0.19314088225364684,
      "num_tokens": 36297120.0,
      "step": 20920
    },
    {
      "entropy": 5.688340520858764,
      "epoch": 1.6280490177008364,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047343489742567593,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20170427858829498,
      "num_tokens": 36305232.0,
      "step": 20925
    },
    {
      "entropy": 5.630097532272339,
      "epoch": 1.628438047072554,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047342183148485424,
      "loss": 5.1254,
      "mean_token_accuracy": 0.2028343990445137,
      "num_tokens": 36313587.0,
      "step": 20930
    },
    {
      "entropy": 5.585933637619019,
      "epoch": 1.6288270764442716,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047340876253329706,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20761501789093018,
      "num_tokens": 36322054.0,
      "step": 20935
    },
    {
      "entropy": 5.661096477508545,
      "epoch": 1.629216105815989,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047339569057120275,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20296951234340668,
      "num_tokens": 36330226.0,
      "step": 20940
    },
    {
      "entropy": 5.613459825515747,
      "epoch": 1.6296051351877066,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047338261559876966,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19479045271873474,
      "num_tokens": 36338436.0,
      "step": 20945
    },
    {
      "entropy": 5.562045764923096,
      "epoch": 1.6299941645594243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047336953761619604,
      "loss": 5.0301,
      "mean_token_accuracy": 0.19537149667739867,
      "num_tokens": 36347204.0,
      "step": 20950
    },
    {
      "entropy": 5.619100427627563,
      "epoch": 1.630383193931142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047335645662368046,
      "loss": 5.0394,
      "mean_token_accuracy": 0.2013746440410614,
      "num_tokens": 36356225.0,
      "step": 20955
    },
    {
      "entropy": 5.64088077545166,
      "epoch": 1.6307722233028592,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004733433726214214,
      "loss": 5.085,
      "mean_token_accuracy": 0.19434978663921357,
      "num_tokens": 36364526.0,
      "step": 20960
    },
    {
      "entropy": 5.6659472465515135,
      "epoch": 1.631161252674577,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047333028560961733,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1919610768556595,
      "num_tokens": 36373162.0,
      "step": 20965
    },
    {
      "entropy": 5.6804241180419925,
      "epoch": 1.6315502820462946,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047331719558846687,
      "loss": 5.0692,
      "mean_token_accuracy": 0.19170292019844054,
      "num_tokens": 36381545.0,
      "step": 20970
    },
    {
      "entropy": 5.633826637268067,
      "epoch": 1.631939311418012,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047330410255816863,
      "loss": 5.0279,
      "mean_token_accuracy": 0.20188958197832108,
      "num_tokens": 36390267.0,
      "step": 20975
    },
    {
      "entropy": 5.628381252288818,
      "epoch": 1.6323283407897295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047329100651892123,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2020599976181984,
      "num_tokens": 36398910.0,
      "step": 20980
    },
    {
      "entropy": 5.651256513595581,
      "epoch": 1.6327173701614472,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004732779074709234,
      "loss": 5.0375,
      "mean_token_accuracy": 0.2019544303417206,
      "num_tokens": 36407382.0,
      "step": 20985
    },
    {
      "entropy": 5.614229536056518,
      "epoch": 1.6331063995331647,
      "grad_norm": 1.265625,
      "learning_rate": 0.000473264805414374,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20024768114089966,
      "num_tokens": 36415734.0,
      "step": 20990
    },
    {
      "entropy": 5.660815763473511,
      "epoch": 1.6334954289048822,
      "grad_norm": 1.125,
      "learning_rate": 0.00047325170034947167,
      "loss": 5.1174,
      "mean_token_accuracy": 0.18745462745428085,
      "num_tokens": 36424537.0,
      "step": 20995
    },
    {
      "entropy": 5.613435554504394,
      "epoch": 1.6338844582766,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047323859227641537,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20331979990005494,
      "num_tokens": 36433276.0,
      "step": 21000
    },
    {
      "epoch": 1.6338844582766,
      "eval_entropy": 5.434298949659272,
      "eval_loss": 5.126439094543457,
      "eval_mean_token_accuracy": 0.20373761542464128,
      "eval_num_tokens": 36433276.0,
      "eval_runtime": 21.8698,
      "eval_samples_per_second": 1900.242,
      "eval_steps_per_second": 237.542,
      "step": 21000
    },
    {
      "entropy": 5.668826770782471,
      "epoch": 1.6342734876483176,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047322548119540397,
      "loss": 5.0848,
      "mean_token_accuracy": 0.1948118105530739,
      "num_tokens": 36442154.0,
      "step": 21005
    },
    {
      "entropy": 5.590239667892456,
      "epoch": 1.634662517020035,
      "grad_norm": 1.125,
      "learning_rate": 0.00047321236710663636,
      "loss": 4.9302,
      "mean_token_accuracy": 0.20378666520118713,
      "num_tokens": 36450625.0,
      "step": 21010
    },
    {
      "entropy": 5.654635715484619,
      "epoch": 1.6350515463917525,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047319925001031165,
      "loss": 5.0574,
      "mean_token_accuracy": 0.19998226910829545,
      "num_tokens": 36460082.0,
      "step": 21015
    },
    {
      "entropy": 5.665618848800659,
      "epoch": 1.6354405757634702,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004731861299066287,
      "loss": 5.113,
      "mean_token_accuracy": 0.1916134014725685,
      "num_tokens": 36469372.0,
      "step": 21020
    },
    {
      "entropy": 5.609025669097901,
      "epoch": 1.6358296051351877,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004731730067957867,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2012546494603157,
      "num_tokens": 36478190.0,
      "step": 21025
    },
    {
      "entropy": 5.578097677230835,
      "epoch": 1.6362186345069052,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047315988067798476,
      "loss": 5.038,
      "mean_token_accuracy": 0.2016855552792549,
      "num_tokens": 36486245.0,
      "step": 21030
    },
    {
      "entropy": 5.607224655151367,
      "epoch": 1.6366076638786229,
      "grad_norm": 1.0625,
      "learning_rate": 0.000473146751553422,
      "loss": 5.0849,
      "mean_token_accuracy": 0.1973696082830429,
      "num_tokens": 36495858.0,
      "step": 21035
    },
    {
      "entropy": 5.69123215675354,
      "epoch": 1.6369966932503404,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004731336194222978,
      "loss": 5.0829,
      "mean_token_accuracy": 0.1954067423939705,
      "num_tokens": 36504564.0,
      "step": 21040
    },
    {
      "entropy": 5.554825401306152,
      "epoch": 1.6373857226220578,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047312048428481114,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2021658793091774,
      "num_tokens": 36512758.0,
      "step": 21045
    },
    {
      "entropy": 5.550291252136231,
      "epoch": 1.6377747519937755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047310734614116153,
      "loss": 4.9451,
      "mean_token_accuracy": 0.20209022760391235,
      "num_tokens": 36521429.0,
      "step": 21050
    },
    {
      "entropy": 5.63755578994751,
      "epoch": 1.6381637813654932,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004730942049915483,
      "loss": 5.0371,
      "mean_token_accuracy": 0.20289296954870223,
      "num_tokens": 36530161.0,
      "step": 21055
    },
    {
      "entropy": 5.650422620773315,
      "epoch": 1.6385528107372107,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004730810608361707,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19996340274810792,
      "num_tokens": 36538184.0,
      "step": 21060
    },
    {
      "entropy": 5.601419448852539,
      "epoch": 1.6389418401089282,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047306791367522833,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2049474149942398,
      "num_tokens": 36547418.0,
      "step": 21065
    },
    {
      "entropy": 5.750905132293701,
      "epoch": 1.6393308694806459,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004730547635089206,
      "loss": 5.1935,
      "mean_token_accuracy": 0.19259399473667144,
      "num_tokens": 36556542.0,
      "step": 21070
    },
    {
      "entropy": 5.650908946990967,
      "epoch": 1.6397198988523634,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047304161033744714,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20329208821058273,
      "num_tokens": 36565903.0,
      "step": 21075
    },
    {
      "entropy": 5.578463792800903,
      "epoch": 1.6401089282240808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004730284541610075,
      "loss": 5.0477,
      "mean_token_accuracy": 0.19732554405927658,
      "num_tokens": 36574010.0,
      "step": 21080
    },
    {
      "entropy": 5.595831441879272,
      "epoch": 1.6404979575957985,
      "grad_norm": 1.125,
      "learning_rate": 0.0004730152949798012,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20015841722488403,
      "num_tokens": 36582142.0,
      "step": 21085
    },
    {
      "entropy": 5.59624171257019,
      "epoch": 1.640886986967516,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047300213279402787,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20062974691390992,
      "num_tokens": 36590676.0,
      "step": 21090
    },
    {
      "entropy": 5.619918441772461,
      "epoch": 1.6412760163392335,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047298896760388745,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21270676106214523,
      "num_tokens": 36599489.0,
      "step": 21095
    },
    {
      "entropy": 5.653407907485962,
      "epoch": 1.6416650457109512,
      "grad_norm": 1.125,
      "learning_rate": 0.0004729757994095796,
      "loss": 5.0239,
      "mean_token_accuracy": 0.19771330654621125,
      "num_tokens": 36608231.0,
      "step": 21100
    },
    {
      "entropy": 5.557680988311768,
      "epoch": 1.6420540750826689,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047296262821130405,
      "loss": 5.1146,
      "mean_token_accuracy": 0.1897520750761032,
      "num_tokens": 36617335.0,
      "step": 21105
    },
    {
      "entropy": 5.691675186157227,
      "epoch": 1.6424431044543863,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047294945400926065,
      "loss": 5.0881,
      "mean_token_accuracy": 0.19582380950450898,
      "num_tokens": 36626249.0,
      "step": 21110
    },
    {
      "entropy": 5.738373422622681,
      "epoch": 1.6428321338261038,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004729362768036494,
      "loss": 5.1374,
      "mean_token_accuracy": 0.19252006560564042,
      "num_tokens": 36634602.0,
      "step": 21115
    },
    {
      "entropy": 5.658088493347168,
      "epoch": 1.6432211631978215,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047292309659467024,
      "loss": 5.082,
      "mean_token_accuracy": 0.19650650024414062,
      "num_tokens": 36643091.0,
      "step": 21120
    },
    {
      "entropy": 5.661387825012207,
      "epoch": 1.643610192569539,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004729099133825231,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19936486929655076,
      "num_tokens": 36652530.0,
      "step": 21125
    },
    {
      "entropy": 5.715387487411499,
      "epoch": 1.6439992219412565,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000472896727167408,
      "loss": 5.1023,
      "mean_token_accuracy": 0.19053196609020234,
      "num_tokens": 36660858.0,
      "step": 21130
    },
    {
      "entropy": 5.76104154586792,
      "epoch": 1.6443882513129742,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047288353794952505,
      "loss": 5.1838,
      "mean_token_accuracy": 0.185442516207695,
      "num_tokens": 36670211.0,
      "step": 21135
    },
    {
      "entropy": 5.642037868499756,
      "epoch": 1.6447772806846916,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004728703457290744,
      "loss": 5.095,
      "mean_token_accuracy": 0.19374112635850907,
      "num_tokens": 36679604.0,
      "step": 21140
    },
    {
      "entropy": 5.673187398910523,
      "epoch": 1.6451663100564091,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004728571505062562,
      "loss": 5.1658,
      "mean_token_accuracy": 0.18581098914146424,
      "num_tokens": 36687908.0,
      "step": 21145
    },
    {
      "entropy": 5.567936801910401,
      "epoch": 1.6455553394281268,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004728439522812707,
      "loss": 4.968,
      "mean_token_accuracy": 0.20611024349927903,
      "num_tokens": 36696675.0,
      "step": 21150
    },
    {
      "entropy": 5.6156738758087155,
      "epoch": 1.6459443687998445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047283075105431806,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19575029611587524,
      "num_tokens": 36705594.0,
      "step": 21155
    },
    {
      "entropy": 5.740062522888183,
      "epoch": 1.646333398171562,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004728175468255987,
      "loss": 5.091,
      "mean_token_accuracy": 0.19588096737861632,
      "num_tokens": 36714008.0,
      "step": 21160
    },
    {
      "entropy": 5.609795379638672,
      "epoch": 1.6467224275432795,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047280433959531287,
      "loss": 4.9566,
      "mean_token_accuracy": 0.20459319204092025,
      "num_tokens": 36723093.0,
      "step": 21165
    },
    {
      "entropy": 5.59610276222229,
      "epoch": 1.6471114569149972,
      "grad_norm": 1.09375,
      "learning_rate": 0.000472791129363661,
      "loss": 4.9443,
      "mean_token_accuracy": 0.20283858329057694,
      "num_tokens": 36732119.0,
      "step": 21170
    },
    {
      "entropy": 5.5894612789154055,
      "epoch": 1.6475004862867146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004727779161308436,
      "loss": 5.0549,
      "mean_token_accuracy": 0.19333341717720032,
      "num_tokens": 36740593.0,
      "step": 21175
    },
    {
      "entropy": 5.718648195266724,
      "epoch": 1.6478895156584321,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004727646998970612,
      "loss": 5.2164,
      "mean_token_accuracy": 0.18926844298839568,
      "num_tokens": 36749083.0,
      "step": 21180
    },
    {
      "entropy": 5.680900239944458,
      "epoch": 1.6482785450301498,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047275148066251417,
      "loss": 5.029,
      "mean_token_accuracy": 0.1985544115304947,
      "num_tokens": 36757469.0,
      "step": 21185
    },
    {
      "entropy": 5.707893991470337,
      "epoch": 1.6486675744018673,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004727382584274032,
      "loss": 5.1665,
      "mean_token_accuracy": 0.19282641410827636,
      "num_tokens": 36766197.0,
      "step": 21190
    },
    {
      "entropy": 5.60744571685791,
      "epoch": 1.6490566037735848,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047272503319192887,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20094352066516877,
      "num_tokens": 36774172.0,
      "step": 21195
    },
    {
      "entropy": 5.64321985244751,
      "epoch": 1.6494456331453025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004727118049562919,
      "loss": 5.0788,
      "mean_token_accuracy": 0.19193458706140518,
      "num_tokens": 36783549.0,
      "step": 21200
    },
    {
      "entropy": 5.758845615386963,
      "epoch": 1.6498346625170202,
      "grad_norm": 1.140625,
      "learning_rate": 0.000472698573720693,
      "loss": 5.1255,
      "mean_token_accuracy": 0.19129636138677597,
      "num_tokens": 36791755.0,
      "step": 21205
    },
    {
      "entropy": 5.687686538696289,
      "epoch": 1.6502236918887376,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004726853394853329,
      "loss": 5.1706,
      "mean_token_accuracy": 0.18552551716566085,
      "num_tokens": 36801253.0,
      "step": 21210
    },
    {
      "entropy": 5.600636148452759,
      "epoch": 1.650612721260455,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047267210225041234,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19830249845981598,
      "num_tokens": 36810497.0,
      "step": 21215
    },
    {
      "entropy": 5.607898378372193,
      "epoch": 1.6510017506321728,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047265886201613234,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20748896449804305,
      "num_tokens": 36819823.0,
      "step": 21220
    },
    {
      "entropy": 5.629617929458618,
      "epoch": 1.6513907800038903,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047264561878269365,
      "loss": 4.9974,
      "mean_token_accuracy": 0.19805109798908233,
      "num_tokens": 36828935.0,
      "step": 21225
    },
    {
      "entropy": 5.603018522262573,
      "epoch": 1.6517798093756078,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004726323725502973,
      "loss": 4.9483,
      "mean_token_accuracy": 0.20436916798353194,
      "num_tokens": 36837906.0,
      "step": 21230
    },
    {
      "entropy": 5.644351673126221,
      "epoch": 1.6521688387473255,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004726191233191443,
      "loss": 5.035,
      "mean_token_accuracy": 0.20484526306390763,
      "num_tokens": 36846041.0,
      "step": 21235
    },
    {
      "entropy": 5.676253318786621,
      "epoch": 1.6525578681190431,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726058710894357,
      "loss": 5.1177,
      "mean_token_accuracy": 0.1943038210272789,
      "num_tokens": 36854493.0,
      "step": 21240
    },
    {
      "entropy": 5.588175868988037,
      "epoch": 1.6529468974907604,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004725926158613724,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2043202042579651,
      "num_tokens": 36863042.0,
      "step": 21245
    },
    {
      "entropy": 5.688817358016967,
      "epoch": 1.653335926862478,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004725793576351557,
      "loss": 5.0936,
      "mean_token_accuracy": 0.1938068225979805,
      "num_tokens": 36872513.0,
      "step": 21250
    },
    {
      "entropy": 5.683014345169068,
      "epoch": 1.6537249562341958,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004725660964109867,
      "loss": 5.0988,
      "mean_token_accuracy": 0.19154474884271622,
      "num_tokens": 36881149.0,
      "step": 21255
    },
    {
      "entropy": 5.651687049865723,
      "epoch": 1.6541139856059133,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047255283218906673,
      "loss": 5.0971,
      "mean_token_accuracy": 0.20226135104894638,
      "num_tokens": 36889372.0,
      "step": 21260
    },
    {
      "entropy": 5.592989063262939,
      "epoch": 1.6545030149776307,
      "grad_norm": 1.171875,
      "learning_rate": 0.000472539564969597,
      "loss": 5.0494,
      "mean_token_accuracy": 0.19565453380346298,
      "num_tokens": 36897727.0,
      "step": 21265
    },
    {
      "entropy": 5.62372145652771,
      "epoch": 1.6548920443493484,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047252629475277863,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2016211047768593,
      "num_tokens": 36906288.0,
      "step": 21270
    },
    {
      "entropy": 5.644999742507935,
      "epoch": 1.655281073721066,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004725130215388132,
      "loss": 5.1041,
      "mean_token_accuracy": 0.1936965450644493,
      "num_tokens": 36915614.0,
      "step": 21275
    },
    {
      "entropy": 5.624772024154663,
      "epoch": 1.6556701030927834,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004724997453279021,
      "loss": 5.0006,
      "mean_token_accuracy": 0.2040416732430458,
      "num_tokens": 36923631.0,
      "step": 21280
    },
    {
      "entropy": 5.68469500541687,
      "epoch": 1.656059132464501,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004724864661202467,
      "loss": 5.1176,
      "mean_token_accuracy": 0.1970871016383171,
      "num_tokens": 36932296.0,
      "step": 21285
    },
    {
      "entropy": 5.546028804779053,
      "epoch": 1.6564481618362188,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047247318391604846,
      "loss": 4.943,
      "mean_token_accuracy": 0.20916468948125838,
      "num_tokens": 36940785.0,
      "step": 21290
    },
    {
      "entropy": 5.657522964477539,
      "epoch": 1.656837191207936,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047245989871550897,
      "loss": 5.0618,
      "mean_token_accuracy": 0.20072626024484636,
      "num_tokens": 36948797.0,
      "step": 21295
    },
    {
      "entropy": 5.700917768478393,
      "epoch": 1.6572262205796537,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004724466105188299,
      "loss": 5.1397,
      "mean_token_accuracy": 0.1990715518593788,
      "num_tokens": 36957808.0,
      "step": 21300
    },
    {
      "entropy": 5.640323305130005,
      "epoch": 1.6576152499513714,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047243331932621263,
      "loss": 5.0556,
      "mean_token_accuracy": 0.201519376039505,
      "num_tokens": 36966260.0,
      "step": 21305
    },
    {
      "entropy": 5.6648509979248045,
      "epoch": 1.658004279323089,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004724200251378591,
      "loss": 5.1065,
      "mean_token_accuracy": 0.18817266076803207,
      "num_tokens": 36975436.0,
      "step": 21310
    },
    {
      "entropy": 5.732514238357544,
      "epoch": 1.6583933086948064,
      "grad_norm": 1.125,
      "learning_rate": 0.0004724067279539709,
      "loss": 5.178,
      "mean_token_accuracy": 0.19519775211811066,
      "num_tokens": 36983485.0,
      "step": 21315
    },
    {
      "entropy": 5.7634251594543455,
      "epoch": 1.658782338066524,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047239342777474975,
      "loss": 5.1846,
      "mean_token_accuracy": 0.188082255423069,
      "num_tokens": 36992632.0,
      "step": 21320
    },
    {
      "entropy": 5.706624984741211,
      "epoch": 1.6591713674382416,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047238012460039765,
      "loss": 4.985,
      "mean_token_accuracy": 0.20288124531507493,
      "num_tokens": 37000694.0,
      "step": 21325
    },
    {
      "entropy": 5.624342584609986,
      "epoch": 1.659560396809959,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004723668184311162,
      "loss": 5.0552,
      "mean_token_accuracy": 0.19991773068904878,
      "num_tokens": 37009514.0,
      "step": 21330
    },
    {
      "entropy": 5.726153135299683,
      "epoch": 1.6599494261816767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004723535092671074,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19292595833539963,
      "num_tokens": 37019434.0,
      "step": 21335
    },
    {
      "entropy": 5.623427820205689,
      "epoch": 1.6603384555533944,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047234019710857335,
      "loss": 5.0075,
      "mean_token_accuracy": 0.19335345476865767,
      "num_tokens": 37028673.0,
      "step": 21340
    },
    {
      "entropy": 5.596810817718506,
      "epoch": 1.6607274849251117,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004723268819557158,
      "loss": 5.0301,
      "mean_token_accuracy": 0.1971658855676651,
      "num_tokens": 37038204.0,
      "step": 21345
    },
    {
      "entropy": 5.57248010635376,
      "epoch": 1.6611165142968294,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047231356380873687,
      "loss": 4.9704,
      "mean_token_accuracy": 0.19987514168024062,
      "num_tokens": 37047301.0,
      "step": 21350
    },
    {
      "entropy": 5.6412159442901615,
      "epoch": 1.661505543668547,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004723002426678388,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20067333728075026,
      "num_tokens": 37054892.0,
      "step": 21355
    },
    {
      "entropy": 5.620475435256958,
      "epoch": 1.6618945730402646,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004722869185332235,
      "loss": 4.9782,
      "mean_token_accuracy": 0.19753111749887467,
      "num_tokens": 37063139.0,
      "step": 21360
    },
    {
      "entropy": 5.660967397689819,
      "epoch": 1.662283602411982,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047227359140509324,
      "loss": 5.1135,
      "mean_token_accuracy": 0.19368239641189575,
      "num_tokens": 37072019.0,
      "step": 21365
    },
    {
      "entropy": 5.6679572582244875,
      "epoch": 1.6626726317836997,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004722602612836501,
      "loss": 5.1949,
      "mean_token_accuracy": 0.1887849435210228,
      "num_tokens": 37081209.0,
      "step": 21370
    },
    {
      "entropy": 5.588965654373169,
      "epoch": 1.6630616611554172,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047224692816909665,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20945039987564087,
      "num_tokens": 37089464.0,
      "step": 21375
    },
    {
      "entropy": 5.5879889011383055,
      "epoch": 1.6634506905271347,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004722335920616349,
      "loss": 4.9824,
      "mean_token_accuracy": 0.1996256873011589,
      "num_tokens": 37098224.0,
      "step": 21380
    },
    {
      "entropy": 5.571327590942383,
      "epoch": 1.6638397198988524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047222025296146745,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1970488592982292,
      "num_tokens": 37107039.0,
      "step": 21385
    },
    {
      "entropy": 5.661564302444458,
      "epoch": 1.66422874927057,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047220691086879643,
      "loss": 5.1063,
      "mean_token_accuracy": 0.1938042402267456,
      "num_tokens": 37116029.0,
      "step": 21390
    },
    {
      "entropy": 5.570654726028442,
      "epoch": 1.6646177786422873,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004721935657838245,
      "loss": 4.9184,
      "mean_token_accuracy": 0.21028432846069336,
      "num_tokens": 37124553.0,
      "step": 21395
    },
    {
      "entropy": 5.624243307113647,
      "epoch": 1.665006808014005,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000472180217706754,
      "loss": 5.1033,
      "mean_token_accuracy": 0.19629158079624176,
      "num_tokens": 37132670.0,
      "step": 21400
    },
    {
      "entropy": 5.668309545516967,
      "epoch": 1.6653958373857227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004721668666377876,
      "loss": 5.1748,
      "mean_token_accuracy": 0.19265587329864503,
      "num_tokens": 37141534.0,
      "step": 21405
    },
    {
      "entropy": 5.689278221130371,
      "epoch": 1.6657848667574402,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047215351257712783,
      "loss": 5.1003,
      "mean_token_accuracy": 0.19268966615200042,
      "num_tokens": 37150895.0,
      "step": 21410
    },
    {
      "entropy": 5.654538106918335,
      "epoch": 1.6661738961291577,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047214015552497735,
      "loss": 5.0433,
      "mean_token_accuracy": 0.1996357709169388,
      "num_tokens": 37159244.0,
      "step": 21415
    },
    {
      "entropy": 5.594687557220459,
      "epoch": 1.6665629255008754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047212679548153867,
      "loss": 5.1009,
      "mean_token_accuracy": 0.1999282196164131,
      "num_tokens": 37167762.0,
      "step": 21420
    },
    {
      "entropy": 5.553576421737671,
      "epoch": 1.6669519548725928,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047211343244701475,
      "loss": 5.0292,
      "mean_token_accuracy": 0.19588248878717424,
      "num_tokens": 37176470.0,
      "step": 21425
    },
    {
      "entropy": 5.667855501174927,
      "epoch": 1.6673409842443103,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004721000664216081,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19569285064935685,
      "num_tokens": 37185923.0,
      "step": 21430
    },
    {
      "entropy": 5.649548149108886,
      "epoch": 1.667730013616028,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004720866974055219,
      "loss": 5.1266,
      "mean_token_accuracy": 0.19502634704113006,
      "num_tokens": 37195400.0,
      "step": 21435
    },
    {
      "entropy": 5.704450941085815,
      "epoch": 1.6681190429877457,
      "grad_norm": 1.125,
      "learning_rate": 0.0004720733253989585,
      "loss": 5.1789,
      "mean_token_accuracy": 0.1889655813574791,
      "num_tokens": 37203862.0,
      "step": 21440
    },
    {
      "entropy": 5.6003861904144285,
      "epoch": 1.6685080723594632,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047205995040212114,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20909508764743806,
      "num_tokens": 37212518.0,
      "step": 21445
    },
    {
      "entropy": 5.701730060577392,
      "epoch": 1.6688971017311807,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004720465724152127,
      "loss": 5.169,
      "mean_token_accuracy": 0.18298301547765733,
      "num_tokens": 37221342.0,
      "step": 21450
    },
    {
      "entropy": 5.6239207744598385,
      "epoch": 1.6692861311028984,
      "grad_norm": 1.125,
      "learning_rate": 0.00047203319143843613,
      "loss": 5.0854,
      "mean_token_accuracy": 0.19605720341205596,
      "num_tokens": 37230477.0,
      "step": 21455
    },
    {
      "entropy": 5.65423583984375,
      "epoch": 1.6696751604746158,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004720198074719946,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20570616126060487,
      "num_tokens": 37238590.0,
      "step": 21460
    },
    {
      "entropy": 5.663268661499023,
      "epoch": 1.6700641898463333,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047200642051609094,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19657767415046692,
      "num_tokens": 37248065.0,
      "step": 21465
    },
    {
      "entropy": 5.685648965835571,
      "epoch": 1.670453219218051,
      "grad_norm": 1.25,
      "learning_rate": 0.0004719930305709285,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20066509693861007,
      "num_tokens": 37256467.0,
      "step": 21470
    },
    {
      "entropy": 5.647742652893067,
      "epoch": 1.6708422485897685,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047197963763671033,
      "loss": 5.0785,
      "mean_token_accuracy": 0.19415740072727203,
      "num_tokens": 37264255.0,
      "step": 21475
    },
    {
      "entropy": 5.603282833099366,
      "epoch": 1.671231277961486,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004719662417136397,
      "loss": 4.9826,
      "mean_token_accuracy": 0.20190958082675933,
      "num_tokens": 37272671.0,
      "step": 21480
    },
    {
      "entropy": 5.632424068450928,
      "epoch": 1.6716203073332037,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047195284280191987,
      "loss": 5.0728,
      "mean_token_accuracy": 0.2013126090168953,
      "num_tokens": 37281895.0,
      "step": 21485
    },
    {
      "entropy": 5.554993629455566,
      "epoch": 1.6720093367049214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004719394409017541,
      "loss": 4.9892,
      "mean_token_accuracy": 0.208169187605381,
      "num_tokens": 37290098.0,
      "step": 21490
    },
    {
      "entropy": 5.640008020401001,
      "epoch": 1.6723983660766388,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047192603601334585,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2020018771290779,
      "num_tokens": 37299494.0,
      "step": 21495
    },
    {
      "entropy": 5.716394329071045,
      "epoch": 1.6727873954483563,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047191262813689835,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19476645439863205,
      "num_tokens": 37308366.0,
      "step": 21500
    },
    {
      "entropy": 5.680360317230225,
      "epoch": 1.673176424820074,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004718992172726152,
      "loss": 5.0732,
      "mean_token_accuracy": 0.19651864171028138,
      "num_tokens": 37316774.0,
      "step": 21505
    },
    {
      "entropy": 5.680277490615845,
      "epoch": 1.6735654541917915,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047188580342069983,
      "loss": 5.0937,
      "mean_token_accuracy": 0.18967730700969695,
      "num_tokens": 37325388.0,
      "step": 21510
    },
    {
      "entropy": 5.6967527866363525,
      "epoch": 1.673954483563509,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004718723865813557,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19092942625284196,
      "num_tokens": 37333476.0,
      "step": 21515
    },
    {
      "entropy": 5.728287839889527,
      "epoch": 1.6743435129352267,
      "grad_norm": 1.125,
      "learning_rate": 0.0004718589667547865,
      "loss": 5.1696,
      "mean_token_accuracy": 0.1950910970568657,
      "num_tokens": 37342556.0,
      "step": 21520
    },
    {
      "entropy": 5.6935951709747314,
      "epoch": 1.6747325423069441,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004718455439411958,
      "loss": 5.049,
      "mean_token_accuracy": 0.20312439352273942,
      "num_tokens": 37351082.0,
      "step": 21525
    },
    {
      "entropy": 5.677677154541016,
      "epoch": 1.6751215716786616,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004718321181407873,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20039399117231368,
      "num_tokens": 37359083.0,
      "step": 21530
    },
    {
      "entropy": 5.5726031303405765,
      "epoch": 1.6755106010503793,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004718186893537647,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19430970251560212,
      "num_tokens": 37367982.0,
      "step": 21535
    },
    {
      "entropy": 5.691591548919678,
      "epoch": 1.675899630422097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004718052575803318,
      "loss": 5.1165,
      "mean_token_accuracy": 0.1979105919599533,
      "num_tokens": 37376932.0,
      "step": 21540
    },
    {
      "entropy": 5.7062821865081785,
      "epoch": 1.6762886597938145,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004717918228206923,
      "loss": 5.121,
      "mean_token_accuracy": 0.19254273772239686,
      "num_tokens": 37386248.0,
      "step": 21545
    },
    {
      "entropy": 5.703535842895508,
      "epoch": 1.676677689165532,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004717783850750501,
      "loss": 5.0532,
      "mean_token_accuracy": 0.19381871521472932,
      "num_tokens": 37395412.0,
      "step": 21550
    },
    {
      "entropy": 5.626397895812988,
      "epoch": 1.6770667185372496,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004717649443436091,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20331156998872757,
      "num_tokens": 37404332.0,
      "step": 21555
    },
    {
      "entropy": 5.669739246368408,
      "epoch": 1.6774557479089671,
      "grad_norm": 1.125,
      "learning_rate": 0.00047175150062657336,
      "loss": 5.1227,
      "mean_token_accuracy": 0.19822941720485687,
      "num_tokens": 37413365.0,
      "step": 21560
    },
    {
      "entropy": 5.660504484176636,
      "epoch": 1.6778447772806846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047173805392414664,
      "loss": 5.006,
      "mean_token_accuracy": 0.19375781565904618,
      "num_tokens": 37422164.0,
      "step": 21565
    },
    {
      "entropy": 5.667116689682007,
      "epoch": 1.6782338066524023,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004717246042365332,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1969469115138054,
      "num_tokens": 37431092.0,
      "step": 21570
    },
    {
      "entropy": 5.690341377258301,
      "epoch": 1.6786228360241198,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047171115156393695,
      "loss": 5.1687,
      "mean_token_accuracy": 0.18650313913822175,
      "num_tokens": 37440832.0,
      "step": 21575
    },
    {
      "entropy": 5.660755586624146,
      "epoch": 1.6790118653958372,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047169769590656207,
      "loss": 5.0862,
      "mean_token_accuracy": 0.19193080961704254,
      "num_tokens": 37449317.0,
      "step": 21580
    },
    {
      "entropy": 5.707340383529663,
      "epoch": 1.679400894767555,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047168423726461273,
      "loss": 5.1164,
      "mean_token_accuracy": 0.19405405670404435,
      "num_tokens": 37457820.0,
      "step": 21585
    },
    {
      "entropy": 5.630513095855713,
      "epoch": 1.6797899241392726,
      "grad_norm": 1.125,
      "learning_rate": 0.0004716707756382931,
      "loss": 4.9772,
      "mean_token_accuracy": 0.20753260105848312,
      "num_tokens": 37466378.0,
      "step": 21590
    },
    {
      "entropy": 5.680223512649536,
      "epoch": 1.6801789535109901,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047165731102780757,
      "loss": 5.103,
      "mean_token_accuracy": 0.18248747736215593,
      "num_tokens": 37475114.0,
      "step": 21595
    },
    {
      "entropy": 5.655133152008057,
      "epoch": 1.6805679828827076,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047164384343336037,
      "loss": 5.1351,
      "mean_token_accuracy": 0.18184176087379456,
      "num_tokens": 37483898.0,
      "step": 21600
    },
    {
      "entropy": 5.697434902191162,
      "epoch": 1.6809570122544253,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047163037285515583,
      "loss": 5.1185,
      "mean_token_accuracy": 0.19714518934488295,
      "num_tokens": 37492420.0,
      "step": 21605
    },
    {
      "entropy": 5.660989236831665,
      "epoch": 1.6813460416261428,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004716168992933982,
      "loss": 5.0131,
      "mean_token_accuracy": 0.2033129408955574,
      "num_tokens": 37501090.0,
      "step": 21610
    },
    {
      "entropy": 5.663851070404053,
      "epoch": 1.6817350709978602,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047160342274829234,
      "loss": 5.156,
      "mean_token_accuracy": 0.19125580340623854,
      "num_tokens": 37510465.0,
      "step": 21615
    },
    {
      "entropy": 5.638567638397217,
      "epoch": 1.682124100369578,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047158994322004223,
      "loss": 5.0627,
      "mean_token_accuracy": 0.20153876543045043,
      "num_tokens": 37519155.0,
      "step": 21620
    },
    {
      "entropy": 5.653360843658447,
      "epoch": 1.6825131297412956,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004715764607088527,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20085000097751618,
      "num_tokens": 37527134.0,
      "step": 21625
    },
    {
      "entropy": 5.660516166687012,
      "epoch": 1.6829021591130129,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004715629752149283,
      "loss": 5.0764,
      "mean_token_accuracy": 0.19939368963241577,
      "num_tokens": 37536087.0,
      "step": 21630
    },
    {
      "entropy": 5.663288164138794,
      "epoch": 1.6832911884847306,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047154948673847356,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20212020725011826,
      "num_tokens": 37543896.0,
      "step": 21635
    },
    {
      "entropy": 5.612263774871826,
      "epoch": 1.6836802178564483,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004715359952796933,
      "loss": 5.0756,
      "mean_token_accuracy": 0.19768381267786025,
      "num_tokens": 37552571.0,
      "step": 21640
    },
    {
      "entropy": 5.662260484695435,
      "epoch": 1.6840692472281658,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047152250083879203,
      "loss": 5.1019,
      "mean_token_accuracy": 0.1945017471909523,
      "num_tokens": 37560823.0,
      "step": 21645
    },
    {
      "entropy": 5.632277584075927,
      "epoch": 1.6844582765998832,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047150900341597464,
      "loss": 5.0194,
      "mean_token_accuracy": 0.19122157841920853,
      "num_tokens": 37568312.0,
      "step": 21650
    },
    {
      "entropy": 5.6558318614959715,
      "epoch": 1.684847305971601,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004714955030114459,
      "loss": 5.05,
      "mean_token_accuracy": 0.20261992365121842,
      "num_tokens": 37577008.0,
      "step": 21655
    },
    {
      "entropy": 5.666969633102417,
      "epoch": 1.6852363353433184,
      "grad_norm": 1.125,
      "learning_rate": 0.00047148199962541065,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19433175176382064,
      "num_tokens": 37586252.0,
      "step": 21660
    },
    {
      "entropy": 5.663171482086182,
      "epoch": 1.6856253647150359,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004714684932580738,
      "loss": 5.06,
      "mean_token_accuracy": 0.19799047857522964,
      "num_tokens": 37594699.0,
      "step": 21665
    },
    {
      "entropy": 5.611803913116455,
      "epoch": 1.6860143940867536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004714549839096403,
      "loss": 5.0442,
      "mean_token_accuracy": 0.19953975230455398,
      "num_tokens": 37603891.0,
      "step": 21670
    },
    {
      "entropy": 5.628356456756592,
      "epoch": 1.6864034234584713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047144147158031507,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20155830979347228,
      "num_tokens": 37612809.0,
      "step": 21675
    },
    {
      "entropy": 5.6429952621459964,
      "epoch": 1.6867924528301885,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004714279562703032,
      "loss": 5.0022,
      "mean_token_accuracy": 0.19458650350570678,
      "num_tokens": 37621787.0,
      "step": 21680
    },
    {
      "entropy": 5.595140695571899,
      "epoch": 1.6871814822019062,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004714144379798098,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21020353734493255,
      "num_tokens": 37629796.0,
      "step": 21685
    },
    {
      "entropy": 5.675281715393067,
      "epoch": 1.687570511573624,
      "grad_norm": 1.125,
      "learning_rate": 0.00047140091670903986,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20224758982658386,
      "num_tokens": 37638467.0,
      "step": 21690
    },
    {
      "entropy": 5.642069387435913,
      "epoch": 1.6879595409453414,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047138739245819857,
      "loss": 5.0765,
      "mean_token_accuracy": 0.20178669244050979,
      "num_tokens": 37647159.0,
      "step": 21695
    },
    {
      "entropy": 5.6693004131317135,
      "epoch": 1.6883485703170589,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047137386522749124,
      "loss": 5.0583,
      "mean_token_accuracy": 0.19136077910661697,
      "num_tokens": 37655986.0,
      "step": 21700
    },
    {
      "entropy": 5.596974849700928,
      "epoch": 1.6887375996887766,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004713603350171231,
      "loss": 4.9713,
      "mean_token_accuracy": 0.20393931567668916,
      "num_tokens": 37663932.0,
      "step": 21705
    },
    {
      "entropy": 5.687929773330689,
      "epoch": 1.689126629060494,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047134680182729926,
      "loss": 5.1124,
      "mean_token_accuracy": 0.19556780606508256,
      "num_tokens": 37672142.0,
      "step": 21710
    },
    {
      "entropy": 5.66797661781311,
      "epoch": 1.6895156584322115,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004713332656582254,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20090068131685257,
      "num_tokens": 37680569.0,
      "step": 21715
    },
    {
      "entropy": 5.628546237945557,
      "epoch": 1.6899046878039292,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004713197265101066,
      "loss": 5.113,
      "mean_token_accuracy": 0.19068047106266023,
      "num_tokens": 37688852.0,
      "step": 21720
    },
    {
      "entropy": 5.675530910491943,
      "epoch": 1.690293717175647,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047130618438314856,
      "loss": 5.0789,
      "mean_token_accuracy": 0.1933480381965637,
      "num_tokens": 37696900.0,
      "step": 21725
    },
    {
      "entropy": 5.6806227684021,
      "epoch": 1.6906827465473642,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004712926392775565,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19094593375921248,
      "num_tokens": 37706292.0,
      "step": 21730
    },
    {
      "entropy": 5.593683099746704,
      "epoch": 1.6910717759190819,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004712790911935362,
      "loss": 5.0028,
      "mean_token_accuracy": 0.20013055801391602,
      "num_tokens": 37715025.0,
      "step": 21735
    },
    {
      "entropy": 5.69048662185669,
      "epoch": 1.6914608052907996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000471265540131293,
      "loss": 5.1286,
      "mean_token_accuracy": 0.19569131433963777,
      "num_tokens": 37724124.0,
      "step": 21740
    },
    {
      "entropy": 5.658081007003784,
      "epoch": 1.691849834662517,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047125198609103267,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19198536425828933,
      "num_tokens": 37732689.0,
      "step": 21745
    },
    {
      "entropy": 5.659414863586425,
      "epoch": 1.6922388640342345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004712384290729607,
      "loss": 5.0456,
      "mean_token_accuracy": 0.19558746218681336,
      "num_tokens": 37741206.0,
      "step": 21750
    },
    {
      "entropy": 5.647965049743652,
      "epoch": 1.6926278934059522,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000471224869077283,
      "loss": 5.0709,
      "mean_token_accuracy": 0.19708117693662644,
      "num_tokens": 37749549.0,
      "step": 21755
    },
    {
      "entropy": 5.551876354217529,
      "epoch": 1.6930169227776697,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047121130610420527,
      "loss": 4.8643,
      "mean_token_accuracy": 0.21011525243520737,
      "num_tokens": 37758332.0,
      "step": 21760
    },
    {
      "entropy": 5.623547315597534,
      "epoch": 1.6934059521493872,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047119774015393323,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19382032305002211,
      "num_tokens": 37767368.0,
      "step": 21765
    },
    {
      "entropy": 5.664384508132935,
      "epoch": 1.6937949815211049,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047118417122667285,
      "loss": 5.0356,
      "mean_token_accuracy": 0.19860450625419618,
      "num_tokens": 37776426.0,
      "step": 21770
    },
    {
      "entropy": 5.7742620468139645,
      "epoch": 1.6941840108928226,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004711705993226298,
      "loss": 5.1768,
      "mean_token_accuracy": 0.19229977130889891,
      "num_tokens": 37784464.0,
      "step": 21775
    },
    {
      "entropy": 5.5946276664733885,
      "epoch": 1.6945730402645398,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004711570244420102,
      "loss": 5.029,
      "mean_token_accuracy": 0.1915336325764656,
      "num_tokens": 37794307.0,
      "step": 21780
    },
    {
      "entropy": 5.672204685211182,
      "epoch": 1.6949620696362575,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004711434465850199,
      "loss": 5.1091,
      "mean_token_accuracy": 0.1855338454246521,
      "num_tokens": 37803151.0,
      "step": 21785
    },
    {
      "entropy": 5.661850786209106,
      "epoch": 1.6953510990079752,
      "grad_norm": 1.125,
      "learning_rate": 0.0004711298657518651,
      "loss": 5.0855,
      "mean_token_accuracy": 0.1940457582473755,
      "num_tokens": 37812433.0,
      "step": 21790
    },
    {
      "entropy": 5.633480834960937,
      "epoch": 1.6957401283796927,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004711162819427518,
      "loss": 5.0222,
      "mean_token_accuracy": 0.19310323148965836,
      "num_tokens": 37820627.0,
      "step": 21795
    },
    {
      "entropy": 5.632673835754394,
      "epoch": 1.6961291577514102,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000471102695157886,
      "loss": 5.0763,
      "mean_token_accuracy": 0.2001366749405861,
      "num_tokens": 37829015.0,
      "step": 21800
    },
    {
      "entropy": 5.696458101272583,
      "epoch": 1.6965181871231279,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004710891053974739,
      "loss": 5.1528,
      "mean_token_accuracy": 0.19283418655395507,
      "num_tokens": 37837097.0,
      "step": 21805
    },
    {
      "entropy": 5.690842008590698,
      "epoch": 1.6969072164948453,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004710755126617217,
      "loss": 5.0897,
      "mean_token_accuracy": 0.1905737727880478,
      "num_tokens": 37845414.0,
      "step": 21810
    },
    {
      "entropy": 5.687635326385498,
      "epoch": 1.6972962458665628,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004710619169508358,
      "loss": 5.064,
      "mean_token_accuracy": 0.19839192181825638,
      "num_tokens": 37855006.0,
      "step": 21815
    },
    {
      "entropy": 5.770589065551758,
      "epoch": 1.6976852752382805,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004710483182650223,
      "loss": 5.2258,
      "mean_token_accuracy": 0.18923259228467942,
      "num_tokens": 37863611.0,
      "step": 21820
    },
    {
      "entropy": 5.6761058330535885,
      "epoch": 1.6980743046099982,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047103471660448767,
      "loss": 5.1011,
      "mean_token_accuracy": 0.2019271656870842,
      "num_tokens": 37871935.0,
      "step": 21825
    },
    {
      "entropy": 5.631487655639648,
      "epoch": 1.6984633339817157,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047102111196943813,
      "loss": 5.0555,
      "mean_token_accuracy": 0.19711328744888307,
      "num_tokens": 37880760.0,
      "step": 21830
    },
    {
      "entropy": 5.668471956253052,
      "epoch": 1.6988523633534331,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047100750436008035,
      "loss": 5.0016,
      "mean_token_accuracy": 0.19684967398643494,
      "num_tokens": 37889457.0,
      "step": 21835
    },
    {
      "entropy": 5.698253107070923,
      "epoch": 1.6992413927251508,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047099389377662054,
      "loss": 5.0529,
      "mean_token_accuracy": 0.1970360681414604,
      "num_tokens": 37897805.0,
      "step": 21840
    },
    {
      "entropy": 5.674101877212524,
      "epoch": 1.6996304220968683,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004709802802192654,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1962428480386734,
      "num_tokens": 37906292.0,
      "step": 21845
    },
    {
      "entropy": 5.598335552215576,
      "epoch": 1.7000194514685858,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047096666368822153,
      "loss": 5.0351,
      "mean_token_accuracy": 0.19739592522382737,
      "num_tokens": 37914513.0,
      "step": 21850
    },
    {
      "entropy": 5.6805524826049805,
      "epoch": 1.7004084808403035,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047095304418369536,
      "loss": 5.0359,
      "mean_token_accuracy": 0.20112321823835372,
      "num_tokens": 37923097.0,
      "step": 21855
    },
    {
      "entropy": 5.649076128005982,
      "epoch": 1.700797510212021,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004709394217058936,
      "loss": 5.079,
      "mean_token_accuracy": 0.20300059765577316,
      "num_tokens": 37931617.0,
      "step": 21860
    },
    {
      "entropy": 5.652280950546265,
      "epoch": 1.7011865395837384,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004709257962550231,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20259252488613128,
      "num_tokens": 37939777.0,
      "step": 21865
    },
    {
      "entropy": 5.655804014205932,
      "epoch": 1.7015755689554561,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047091216783129035,
      "loss": 5.0681,
      "mean_token_accuracy": 0.19665808379650115,
      "num_tokens": 37949375.0,
      "step": 21870
    },
    {
      "entropy": 5.664904403686523,
      "epoch": 1.7019645983271738,
      "grad_norm": 1.0,
      "learning_rate": 0.0004708985364349024,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20204522907733918,
      "num_tokens": 37958724.0,
      "step": 21875
    },
    {
      "entropy": 5.664713907241821,
      "epoch": 1.7023536276988913,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047088490206606586,
      "loss": 5.023,
      "mean_token_accuracy": 0.20141012966632843,
      "num_tokens": 37967646.0,
      "step": 21880
    },
    {
      "entropy": 5.663529014587402,
      "epoch": 1.7027426570706088,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047087126472498785,
      "loss": 5.103,
      "mean_token_accuracy": 0.19719585180282592,
      "num_tokens": 37976058.0,
      "step": 21885
    },
    {
      "entropy": 5.592619132995606,
      "epoch": 1.7031316864423265,
      "grad_norm": 1.171875,
      "learning_rate": 0.000470857624411875,
      "loss": 4.9978,
      "mean_token_accuracy": 0.20292682945728302,
      "num_tokens": 37984895.0,
      "step": 21890
    },
    {
      "entropy": 5.632656049728394,
      "epoch": 1.703520715814044,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047084398112693456,
      "loss": 5.0021,
      "mean_token_accuracy": 0.20369299799203872,
      "num_tokens": 37993514.0,
      "step": 21895
    },
    {
      "entropy": 5.66341552734375,
      "epoch": 1.7039097451857614,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004708303348703734,
      "loss": 5.041,
      "mean_token_accuracy": 0.19474004507064818,
      "num_tokens": 38002274.0,
      "step": 21900
    },
    {
      "entropy": 5.658758306503296,
      "epoch": 1.7042987745574791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004708166856423986,
      "loss": 4.9745,
      "mean_token_accuracy": 0.19942739307880403,
      "num_tokens": 38010336.0,
      "step": 21905
    },
    {
      "entropy": 5.654409980773925,
      "epoch": 1.7046878039291966,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047080303344321727,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20619186908006668,
      "num_tokens": 38019166.0,
      "step": 21910
    },
    {
      "entropy": 5.60485987663269,
      "epoch": 1.705076833300914,
      "grad_norm": 1.125,
      "learning_rate": 0.00047078937827303653,
      "loss": 5.0787,
      "mean_token_accuracy": 0.19914131611585617,
      "num_tokens": 38027634.0,
      "step": 21915
    },
    {
      "entropy": 5.611303472518921,
      "epoch": 1.7054658626726318,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004707757201320636,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19664876013994217,
      "num_tokens": 38036611.0,
      "step": 21920
    },
    {
      "entropy": 5.583642244338989,
      "epoch": 1.7058548920443495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047076205902050577,
      "loss": 4.9885,
      "mean_token_accuracy": 0.20393583327531814,
      "num_tokens": 38045815.0,
      "step": 21925
    },
    {
      "entropy": 5.681838417053223,
      "epoch": 1.706243921416067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047074839493857024,
      "loss": 5.1378,
      "mean_token_accuracy": 0.18761497288942336,
      "num_tokens": 38054973.0,
      "step": 21930
    },
    {
      "entropy": 5.680505084991455,
      "epoch": 1.7066329507877844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004707347278864644,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20467377007007598,
      "num_tokens": 38064084.0,
      "step": 21935
    },
    {
      "entropy": 5.625653886795044,
      "epoch": 1.7070219801595021,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047072105786439563,
      "loss": 5.0413,
      "mean_token_accuracy": 0.2004789814352989,
      "num_tokens": 38073838.0,
      "step": 21940
    },
    {
      "entropy": 5.7176148891448975,
      "epoch": 1.7074110095312196,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047070738487257137,
      "loss": 5.1356,
      "mean_token_accuracy": 0.18751115202903748,
      "num_tokens": 38082531.0,
      "step": 21945
    },
    {
      "entropy": 5.712042284011841,
      "epoch": 1.707800038902937,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047069370891119895,
      "loss": 5.1394,
      "mean_token_accuracy": 0.18958460539579391,
      "num_tokens": 38091499.0,
      "step": 21950
    },
    {
      "entropy": 5.638915061950684,
      "epoch": 1.7081890682746548,
      "grad_norm": 1.125,
      "learning_rate": 0.000470680029980486,
      "loss": 4.974,
      "mean_token_accuracy": 0.20963735580444337,
      "num_tokens": 38099488.0,
      "step": 21955
    },
    {
      "entropy": 5.666776847839356,
      "epoch": 1.7085780976463723,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047066634808064006,
      "loss": 5.1028,
      "mean_token_accuracy": 0.19414198249578477,
      "num_tokens": 38107535.0,
      "step": 21960
    },
    {
      "entropy": 5.658326148986816,
      "epoch": 1.7089671270180897,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047065266321186873,
      "loss": 5.049,
      "mean_token_accuracy": 0.20387632101774217,
      "num_tokens": 38116457.0,
      "step": 21965
    },
    {
      "entropy": 5.6876777648925785,
      "epoch": 1.7093561563898074,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004706389753743797,
      "loss": 4.9786,
      "mean_token_accuracy": 0.20168186575174332,
      "num_tokens": 38124571.0,
      "step": 21970
    },
    {
      "entropy": 5.615490484237671,
      "epoch": 1.7097451857615251,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004706252845683805,
      "loss": 5.0098,
      "mean_token_accuracy": 0.19724636673927307,
      "num_tokens": 38132789.0,
      "step": 21975
    },
    {
      "entropy": 5.525396680831909,
      "epoch": 1.7101342151332426,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047061159079407903,
      "loss": 4.9339,
      "mean_token_accuracy": 0.20236624479293824,
      "num_tokens": 38141413.0,
      "step": 21980
    },
    {
      "entropy": 5.606761693954468,
      "epoch": 1.71052324450496,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000470597894051683,
      "loss": 4.9808,
      "mean_token_accuracy": 0.196353717148304,
      "num_tokens": 38149869.0,
      "step": 21985
    },
    {
      "entropy": 5.617020320892334,
      "epoch": 1.7109122738766778,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004705841943414003,
      "loss": 4.9074,
      "mean_token_accuracy": 0.20834430009126664,
      "num_tokens": 38157356.0,
      "step": 21990
    },
    {
      "entropy": 5.650063133239746,
      "epoch": 1.7113013032483952,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047057049166343876,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2007967472076416,
      "num_tokens": 38166125.0,
      "step": 21995
    },
    {
      "entropy": 5.619777536392212,
      "epoch": 1.7116903326201127,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047055678601800623,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20387925505638121,
      "num_tokens": 38174171.0,
      "step": 22000
    },
    {
      "entropy": 5.65975513458252,
      "epoch": 1.7120793619918304,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047054307740531076,
      "loss": 5.1367,
      "mean_token_accuracy": 0.1937941625714302,
      "num_tokens": 38182755.0,
      "step": 22005
    },
    {
      "entropy": 5.674558687210083,
      "epoch": 1.712468391363548,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047052936582556035,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20152108669281005,
      "num_tokens": 38191289.0,
      "step": 22010
    },
    {
      "entropy": 5.685760021209717,
      "epoch": 1.7128574207352654,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047051565127896307,
      "loss": 5.178,
      "mean_token_accuracy": 0.18924542367458344,
      "num_tokens": 38200624.0,
      "step": 22015
    },
    {
      "entropy": 5.687732744216919,
      "epoch": 1.713246450106983,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047050193376572686,
      "loss": 5.0057,
      "mean_token_accuracy": 0.19936662018299103,
      "num_tokens": 38209272.0,
      "step": 22020
    },
    {
      "entropy": 5.709579706192017,
      "epoch": 1.7136354794787008,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047048821328606,
      "loss": 5.1651,
      "mean_token_accuracy": 0.1925066202878952,
      "num_tokens": 38218851.0,
      "step": 22025
    },
    {
      "entropy": 5.664924812316895,
      "epoch": 1.7140245088504182,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004704744898401708,
      "loss": 5.0493,
      "mean_token_accuracy": 0.1972934454679489,
      "num_tokens": 38227367.0,
      "step": 22030
    },
    {
      "entropy": 5.6547037124633786,
      "epoch": 1.7144135382221357,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004704607634282672,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20214338451623917,
      "num_tokens": 38235721.0,
      "step": 22035
    },
    {
      "entropy": 5.70013689994812,
      "epoch": 1.7148025675938534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047044703405055765,
      "loss": 5.1148,
      "mean_token_accuracy": 0.190228696167469,
      "num_tokens": 38244686.0,
      "step": 22040
    },
    {
      "entropy": 5.684786605834961,
      "epoch": 1.7151915969655709,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047043330170725046,
      "loss": 5.1327,
      "mean_token_accuracy": 0.19295063316822053,
      "num_tokens": 38253609.0,
      "step": 22045
    },
    {
      "entropy": 5.662104368209839,
      "epoch": 1.7155806263372884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047041956639855406,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20118971467018126,
      "num_tokens": 38261746.0,
      "step": 22050
    },
    {
      "entropy": 5.607313919067383,
      "epoch": 1.715969655709006,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004704058281246766,
      "loss": 5.046,
      "mean_token_accuracy": 0.20168684273958207,
      "num_tokens": 38270573.0,
      "step": 22055
    },
    {
      "entropy": 5.632293844223023,
      "epoch": 1.7163586850807238,
      "grad_norm": 1.25,
      "learning_rate": 0.0004703920868858268,
      "loss": 5.0686,
      "mean_token_accuracy": 0.19501600563526153,
      "num_tokens": 38279031.0,
      "step": 22060
    },
    {
      "entropy": 5.7040746212005615,
      "epoch": 1.716747714452441,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047037834268221315,
      "loss": 5.1131,
      "mean_token_accuracy": 0.19128086268901826,
      "num_tokens": 38287060.0,
      "step": 22065
    },
    {
      "entropy": 5.678301191329956,
      "epoch": 1.7171367438241587,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004703645955140441,
      "loss": 5.0641,
      "mean_token_accuracy": 0.2001593753695488,
      "num_tokens": 38295192.0,
      "step": 22070
    },
    {
      "entropy": 5.612302303314209,
      "epoch": 1.7175257731958764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047035084538152815,
      "loss": 5.0013,
      "mean_token_accuracy": 0.19833966046571733,
      "num_tokens": 38303932.0,
      "step": 22075
    },
    {
      "entropy": 5.701074552536011,
      "epoch": 1.7179148025675939,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004703370922848742,
      "loss": 5.15,
      "mean_token_accuracy": 0.18675490766763686,
      "num_tokens": 38312906.0,
      "step": 22080
    },
    {
      "entropy": 5.668798208236694,
      "epoch": 1.7183038319393114,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047032333622429074,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19245337098836898,
      "num_tokens": 38322359.0,
      "step": 22085
    },
    {
      "entropy": 5.747556543350219,
      "epoch": 1.718692861311029,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047030957719998656,
      "loss": 5.1589,
      "mean_token_accuracy": 0.18749293684959412,
      "num_tokens": 38331880.0,
      "step": 22090
    },
    {
      "entropy": 5.707289695739746,
      "epoch": 1.7190818906827465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004702958152121704,
      "loss": 5.2246,
      "mean_token_accuracy": 0.18746642172336578,
      "num_tokens": 38340763.0,
      "step": 22095
    },
    {
      "entropy": 5.69387845993042,
      "epoch": 1.719470920054464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004702820502610511,
      "loss": 5.0533,
      "mean_token_accuracy": 0.1938890442252159,
      "num_tokens": 38349427.0,
      "step": 22100
    },
    {
      "entropy": 5.648448657989502,
      "epoch": 1.7198599494261817,
      "grad_norm": 1.125,
      "learning_rate": 0.0004702682823468375,
      "loss": 5.0196,
      "mean_token_accuracy": 0.20643892288208007,
      "num_tokens": 38358025.0,
      "step": 22105
    },
    {
      "entropy": 5.655142450332642,
      "epoch": 1.7202489787978994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047025451146973844,
      "loss": 5.019,
      "mean_token_accuracy": 0.19807271808385848,
      "num_tokens": 38366923.0,
      "step": 22110
    },
    {
      "entropy": 5.679611778259277,
      "epoch": 1.7206380081696167,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047024073762996305,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20126565992832185,
      "num_tokens": 38375397.0,
      "step": 22115
    },
    {
      "entropy": 5.644463205337525,
      "epoch": 1.7210270375413343,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004702269608277202,
      "loss": 5.0083,
      "mean_token_accuracy": 0.20640270709991454,
      "num_tokens": 38384279.0,
      "step": 22120
    },
    {
      "entropy": 5.672321510314942,
      "epoch": 1.721416066913052,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004702131810632189,
      "loss": 5.0661,
      "mean_token_accuracy": 0.19860919266939164,
      "num_tokens": 38392087.0,
      "step": 22125
    },
    {
      "entropy": 5.626813173294067,
      "epoch": 1.7218050962847695,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047019939833666837,
      "loss": 5.1564,
      "mean_token_accuracy": 0.19120646566152572,
      "num_tokens": 38400616.0,
      "step": 22130
    },
    {
      "entropy": 5.661971187591552,
      "epoch": 1.722194125656487,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004701856126482776,
      "loss": 5.0513,
      "mean_token_accuracy": 0.1976519912481308,
      "num_tokens": 38408975.0,
      "step": 22135
    },
    {
      "entropy": 5.667482948303222,
      "epoch": 1.7225831550282047,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004701718239982559,
      "loss": 5.1607,
      "mean_token_accuracy": 0.19737916737794875,
      "num_tokens": 38417161.0,
      "step": 22140
    },
    {
      "entropy": 5.695041942596435,
      "epoch": 1.7229721843999222,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047015803238681234,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1965773493051529,
      "num_tokens": 38425078.0,
      "step": 22145
    },
    {
      "entropy": 5.661451911926269,
      "epoch": 1.7233612137716396,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004701442378141563,
      "loss": 5.0224,
      "mean_token_accuracy": 0.19623859822750092,
      "num_tokens": 38433347.0,
      "step": 22150
    },
    {
      "entropy": 5.722309112548828,
      "epoch": 1.7237502431433573,
      "grad_norm": 1.09375,
      "learning_rate": 0.000470130440280497,
      "loss": 5.114,
      "mean_token_accuracy": 0.19073316007852553,
      "num_tokens": 38442373.0,
      "step": 22155
    },
    {
      "entropy": 5.679425954818726,
      "epoch": 1.724139272515075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004701166397860439,
      "loss": 5.1863,
      "mean_token_accuracy": 0.18589166700839996,
      "num_tokens": 38452325.0,
      "step": 22160
    },
    {
      "entropy": 5.741942453384399,
      "epoch": 1.7245283018867923,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004701028363310064,
      "loss": 5.1782,
      "mean_token_accuracy": 0.18969137519598006,
      "num_tokens": 38462317.0,
      "step": 22165
    },
    {
      "entropy": 5.669379043579101,
      "epoch": 1.72491733125851,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047008902991559385,
      "loss": 5.0107,
      "mean_token_accuracy": 0.19835032075643538,
      "num_tokens": 38471149.0,
      "step": 22170
    },
    {
      "entropy": 5.743492364883423,
      "epoch": 1.7253063606302277,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004700752205400158,
      "loss": 5.1174,
      "mean_token_accuracy": 0.19219594597816467,
      "num_tokens": 38480427.0,
      "step": 22175
    },
    {
      "entropy": 5.609868335723877,
      "epoch": 1.7256953900019452,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047006140820448174,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20142827332019805,
      "num_tokens": 38489283.0,
      "step": 22180
    },
    {
      "entropy": 5.655654859542847,
      "epoch": 1.7260844193736626,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047004759290920136,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1923058196902275,
      "num_tokens": 38497557.0,
      "step": 22185
    },
    {
      "entropy": 5.663022994995117,
      "epoch": 1.7264734487453803,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004700337746543841,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20584206730127336,
      "num_tokens": 38506589.0,
      "step": 22190
    },
    {
      "entropy": 5.684738206863403,
      "epoch": 1.7268624781170978,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004700199534402398,
      "loss": 5.0851,
      "mean_token_accuracy": 0.1928133100271225,
      "num_tokens": 38515016.0,
      "step": 22195
    },
    {
      "entropy": 5.740749311447144,
      "epoch": 1.7272515074888153,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004700061292669781,
      "loss": 5.124,
      "mean_token_accuracy": 0.1928827852010727,
      "num_tokens": 38523792.0,
      "step": 22200
    },
    {
      "entropy": 5.545234155654907,
      "epoch": 1.727640536860533,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004699923021348088,
      "loss": 4.9527,
      "mean_token_accuracy": 0.20264324843883513,
      "num_tokens": 38532007.0,
      "step": 22205
    },
    {
      "entropy": 5.594194078445435,
      "epoch": 1.7280295662322507,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046997847204394166,
      "loss": 5.0442,
      "mean_token_accuracy": 0.1982519045472145,
      "num_tokens": 38541044.0,
      "step": 22210
    },
    {
      "entropy": 5.6463038444519045,
      "epoch": 1.728418595603968,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046996463899458653,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2060134544968605,
      "num_tokens": 38549365.0,
      "step": 22215
    },
    {
      "entropy": 5.63493070602417,
      "epoch": 1.7288076249756856,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004699508029869533,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20558922439813615,
      "num_tokens": 38558289.0,
      "step": 22220
    },
    {
      "entropy": 5.698814535140992,
      "epoch": 1.7291966543474033,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046993696402125205,
      "loss": 5.1136,
      "mean_token_accuracy": 0.18840648829936982,
      "num_tokens": 38568422.0,
      "step": 22225
    },
    {
      "entropy": 5.668754625320434,
      "epoch": 1.7295856837191208,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004699231220976925,
      "loss": 5.055,
      "mean_token_accuracy": 0.1978025868535042,
      "num_tokens": 38577009.0,
      "step": 22230
    },
    {
      "entropy": 5.617860651016235,
      "epoch": 1.7299747130908383,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004699092772164849,
      "loss": 4.9302,
      "mean_token_accuracy": 0.21121170073747636,
      "num_tokens": 38584920.0,
      "step": 22235
    },
    {
      "entropy": 5.514134454727173,
      "epoch": 1.730363742462556,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004698954293778392,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20450992733240128,
      "num_tokens": 38593533.0,
      "step": 22240
    },
    {
      "entropy": 5.577052783966065,
      "epoch": 1.7307527718342735,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046988157858196555,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20209478288888932,
      "num_tokens": 38602300.0,
      "step": 22245
    },
    {
      "entropy": 5.651813554763794,
      "epoch": 1.731141801205991,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046986772482907426,
      "loss": 5.0903,
      "mean_token_accuracy": 0.1968725800514221,
      "num_tokens": 38610729.0,
      "step": 22250
    },
    {
      "entropy": 5.641043090820313,
      "epoch": 1.7315308305777086,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004698538681193754,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20745801627635957,
      "num_tokens": 38618192.0,
      "step": 22255
    },
    {
      "entropy": 5.628970050811768,
      "epoch": 1.7319198599494263,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046984000845307907,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20359894633293152,
      "num_tokens": 38627030.0,
      "step": 22260
    },
    {
      "entropy": 5.68828444480896,
      "epoch": 1.7323088893211438,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046982614583039584,
      "loss": 5.1173,
      "mean_token_accuracy": 0.19621069878339767,
      "num_tokens": 38635761.0,
      "step": 22265
    },
    {
      "entropy": 5.617671060562134,
      "epoch": 1.7326979186928613,
      "grad_norm": 1.25,
      "learning_rate": 0.0004698122802515359,
      "loss": 5.0364,
      "mean_token_accuracy": 0.20187563747167586,
      "num_tokens": 38643952.0,
      "step": 22270
    },
    {
      "entropy": 5.640407514572144,
      "epoch": 1.733086948064579,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004697984117167097,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20475339740514756,
      "num_tokens": 38651888.0,
      "step": 22275
    },
    {
      "entropy": 5.659044408798218,
      "epoch": 1.7334759774362964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046978454022612767,
      "loss": 5.1216,
      "mean_token_accuracy": 0.1941988617181778,
      "num_tokens": 38660466.0,
      "step": 22280
    },
    {
      "entropy": 5.699975252151489,
      "epoch": 1.733865006808014,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046977066578000025,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19183725863695145,
      "num_tokens": 38668429.0,
      "step": 22285
    },
    {
      "entropy": 5.680371189117432,
      "epoch": 1.7342540361797316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046975678837853806,
      "loss": 5.1011,
      "mean_token_accuracy": 0.19901301413774491,
      "num_tokens": 38676391.0,
      "step": 22290
    },
    {
      "entropy": 5.707980394363403,
      "epoch": 1.734643065551449,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046974290802195156,
      "loss": 5.1447,
      "mean_token_accuracy": 0.19748384654521942,
      "num_tokens": 38684662.0,
      "step": 22295
    },
    {
      "entropy": 5.667836856842041,
      "epoch": 1.7350320949231666,
      "grad_norm": 1.25,
      "learning_rate": 0.0004697290247104513,
      "loss": 5.0396,
      "mean_token_accuracy": 0.19850922673940657,
      "num_tokens": 38693497.0,
      "step": 22300
    },
    {
      "entropy": 5.609786748886108,
      "epoch": 1.7354211242948843,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046971513844424814,
      "loss": 4.9824,
      "mean_token_accuracy": 0.19847204834222792,
      "num_tokens": 38701898.0,
      "step": 22305
    },
    {
      "entropy": 5.6908042430877686,
      "epoch": 1.735810153666602,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046970124922355265,
      "loss": 5.084,
      "mean_token_accuracy": 0.19396658539772033,
      "num_tokens": 38710910.0,
      "step": 22310
    },
    {
      "entropy": 5.723724031448365,
      "epoch": 1.7361991830383194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004696873570485756,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19737848192453383,
      "num_tokens": 38719612.0,
      "step": 22315
    },
    {
      "entropy": 5.555810213088989,
      "epoch": 1.736588212410037,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004696734619195278,
      "loss": 5.1011,
      "mean_token_accuracy": 0.18984420448541642,
      "num_tokens": 38728332.0,
      "step": 22320
    },
    {
      "entropy": 5.63932957649231,
      "epoch": 1.7369772417817546,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046965956383662,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1951083078980446,
      "num_tokens": 38736830.0,
      "step": 22325
    },
    {
      "entropy": 5.660230922698974,
      "epoch": 1.737366271153472,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004696456628000632,
      "loss": 5.051,
      "mean_token_accuracy": 0.20213656425476073,
      "num_tokens": 38745972.0,
      "step": 22330
    },
    {
      "entropy": 5.58563642501831,
      "epoch": 1.7377553005251896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046963175881006825,
      "loss": 4.9809,
      "mean_token_accuracy": 0.19983863681554795,
      "num_tokens": 38754287.0,
      "step": 22335
    },
    {
      "entropy": 5.590136480331421,
      "epoch": 1.7381443298969073,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004696178518668462,
      "loss": 5.0503,
      "mean_token_accuracy": 0.19850851744413375,
      "num_tokens": 38763212.0,
      "step": 22340
    },
    {
      "entropy": 5.594533681869507,
      "epoch": 1.7385333592686247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046960394197060804,
      "loss": 4.9031,
      "mean_token_accuracy": 0.20806840658187867,
      "num_tokens": 38771668.0,
      "step": 22345
    },
    {
      "entropy": 5.697371530532837,
      "epoch": 1.7389223886403422,
      "grad_norm": 1.125,
      "learning_rate": 0.00046959002912156473,
      "loss": 5.1027,
      "mean_token_accuracy": 0.1935950830578804,
      "num_tokens": 38781184.0,
      "step": 22350
    },
    {
      "entropy": 5.649533271789551,
      "epoch": 1.73931141801206,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004695761133199275,
      "loss": 4.9374,
      "mean_token_accuracy": 0.20961211919784545,
      "num_tokens": 38789622.0,
      "step": 22355
    },
    {
      "entropy": 5.641437101364136,
      "epoch": 1.7397004473837776,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004695621945659074,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20018697232007981,
      "num_tokens": 38798250.0,
      "step": 22360
    },
    {
      "entropy": 5.672954177856445,
      "epoch": 1.740089476755495,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004695482728597157,
      "loss": 5.055,
      "mean_token_accuracy": 0.2002064824104309,
      "num_tokens": 38807724.0,
      "step": 22365
    },
    {
      "entropy": 5.6568409442901615,
      "epoch": 1.7404785061272126,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046953434820156363,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20160023868083954,
      "num_tokens": 38816017.0,
      "step": 22370
    },
    {
      "entropy": 5.642955684661866,
      "epoch": 1.7408675354989303,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004695204205916624,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20596525371074675,
      "num_tokens": 38824976.0,
      "step": 22375
    },
    {
      "entropy": 5.605226850509643,
      "epoch": 1.7412565648706477,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004695064900302235,
      "loss": 5.0263,
      "mean_token_accuracy": 0.19878586530685424,
      "num_tokens": 38834358.0,
      "step": 22380
    },
    {
      "entropy": 5.664808940887451,
      "epoch": 1.7416455942423652,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004694925565174581,
      "loss": 5.0939,
      "mean_token_accuracy": 0.19231285005807877,
      "num_tokens": 38842801.0,
      "step": 22385
    },
    {
      "entropy": 5.665209722518921,
      "epoch": 1.742034623614083,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046947862005357777,
      "loss": 5.1182,
      "mean_token_accuracy": 0.19022908210754394,
      "num_tokens": 38851335.0,
      "step": 22390
    },
    {
      "entropy": 5.710493516921997,
      "epoch": 1.7424236529858004,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004694646806387939,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19175349920988083,
      "num_tokens": 38860074.0,
      "step": 22395
    },
    {
      "entropy": 5.687563848495484,
      "epoch": 1.7428126823575179,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004694507382733181,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20445260107517244,
      "num_tokens": 38868319.0,
      "step": 22400
    },
    {
      "entropy": 5.654357814788819,
      "epoch": 1.7432017117292355,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004694367929573618,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20206646919250487,
      "num_tokens": 38876971.0,
      "step": 22405
    },
    {
      "entropy": 5.689477014541626,
      "epoch": 1.7435907411009532,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004694228446911367,
      "loss": 5.1933,
      "mean_token_accuracy": 0.18560272455215454,
      "num_tokens": 38886032.0,
      "step": 22410
    },
    {
      "entropy": 5.71575231552124,
      "epoch": 1.7439797704726707,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004694088934748542,
      "loss": 5.0724,
      "mean_token_accuracy": 0.19499998390674592,
      "num_tokens": 38894505.0,
      "step": 22415
    },
    {
      "entropy": 5.602770185470581,
      "epoch": 1.7443687998443882,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004693949393087263,
      "loss": 4.9625,
      "mean_token_accuracy": 0.20709607601165772,
      "num_tokens": 38903450.0,
      "step": 22420
    },
    {
      "entropy": 5.6399445056915285,
      "epoch": 1.744757829216106,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004693809821929647,
      "loss": 5.048,
      "mean_token_accuracy": 0.19909238964319229,
      "num_tokens": 38911326.0,
      "step": 22425
    },
    {
      "entropy": 5.648026180267334,
      "epoch": 1.7451468585878234,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000469367022127781,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20373102128505707,
      "num_tokens": 38920035.0,
      "step": 22430
    },
    {
      "entropy": 5.721333742141724,
      "epoch": 1.7455358879595408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046935305911338703,
      "loss": 5.0647,
      "mean_token_accuracy": 0.20362452119588853,
      "num_tokens": 38928542.0,
      "step": 22435
    },
    {
      "entropy": 5.678221607208252,
      "epoch": 1.7459249173312585,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004693390931499948,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20546302646398545,
      "num_tokens": 38937486.0,
      "step": 22440
    },
    {
      "entropy": 5.610051536560059,
      "epoch": 1.7463139467029762,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004693251242378162,
      "loss": 5.006,
      "mean_token_accuracy": 0.200470532476902,
      "num_tokens": 38946171.0,
      "step": 22445
    },
    {
      "entropy": 5.611274719238281,
      "epoch": 1.7467029760746935,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046931115237706304,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20277661979198455,
      "num_tokens": 38954214.0,
      "step": 22450
    },
    {
      "entropy": 5.721747255325317,
      "epoch": 1.7470920054464112,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004692971775679475,
      "loss": 5.1067,
      "mean_token_accuracy": 0.19866703748703002,
      "num_tokens": 38962102.0,
      "step": 22455
    },
    {
      "entropy": 5.65501389503479,
      "epoch": 1.7474810348181289,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046928319981068154,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2086581841111183,
      "num_tokens": 38970472.0,
      "step": 22460
    },
    {
      "entropy": 5.635138416290284,
      "epoch": 1.7478700641898464,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004692692191054773,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20064300149679185,
      "num_tokens": 38979151.0,
      "step": 22465
    },
    {
      "entropy": 5.659950160980225,
      "epoch": 1.7482590935615638,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004692552354525468,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2036854386329651,
      "num_tokens": 38987044.0,
      "step": 22470
    },
    {
      "entropy": 5.677900648117065,
      "epoch": 1.7486481229332815,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004692412488521023,
      "loss": 5.1283,
      "mean_token_accuracy": 0.19334679394960402,
      "num_tokens": 38995882.0,
      "step": 22475
    },
    {
      "entropy": 5.661788606643677,
      "epoch": 1.749037152304999,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000469227259304356,
      "loss": 5.1113,
      "mean_token_accuracy": 0.1951193168759346,
      "num_tokens": 39004456.0,
      "step": 22480
    },
    {
      "entropy": 5.703683090209961,
      "epoch": 1.7494261816767165,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046921326680952023,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19685509651899338,
      "num_tokens": 39013363.0,
      "step": 22485
    },
    {
      "entropy": 5.672262144088745,
      "epoch": 1.7498152110484342,
      "grad_norm": 1.125,
      "learning_rate": 0.0004691992713678072,
      "loss": 5.0372,
      "mean_token_accuracy": 0.1959453970193863,
      "num_tokens": 39022116.0,
      "step": 22490
    },
    {
      "entropy": 5.699271965026855,
      "epoch": 1.7502042404201519,
      "grad_norm": 1.125,
      "learning_rate": 0.0004691852729794293,
      "loss": 5.0809,
      "mean_token_accuracy": 0.19595789313316345,
      "num_tokens": 39030737.0,
      "step": 22495
    },
    {
      "entropy": 5.594301557540893,
      "epoch": 1.7505932697918691,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004691712716445991,
      "loss": 5.0053,
      "mean_token_accuracy": 0.2094651237130165,
      "num_tokens": 39039511.0,
      "step": 22500
    },
    {
      "entropy": 5.614451217651367,
      "epoch": 1.7509822991635868,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004691572673635288,
      "loss": 5.005,
      "mean_token_accuracy": 0.20285391062498093,
      "num_tokens": 39048745.0,
      "step": 22505
    },
    {
      "entropy": 5.666562843322754,
      "epoch": 1.7513713285353045,
      "grad_norm": 1.171875,
      "learning_rate": 0.000469143260136431,
      "loss": 5.0708,
      "mean_token_accuracy": 0.19720252454280854,
      "num_tokens": 39057311.0,
      "step": 22510
    },
    {
      "entropy": 5.6998292922973635,
      "epoch": 1.751760357907022,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004691292499635183,
      "loss": 5.0652,
      "mean_token_accuracy": 0.1997191295027733,
      "num_tokens": 39065630.0,
      "step": 22515
    },
    {
      "entropy": 5.617317008972168,
      "epoch": 1.7521493872787395,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004691152368450032,
      "loss": 5.052,
      "mean_token_accuracy": 0.1998201236128807,
      "num_tokens": 39074284.0,
      "step": 22520
    },
    {
      "entropy": 5.654884958267212,
      "epoch": 1.7525384166504572,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046910122078109835,
      "loss": 5.0816,
      "mean_token_accuracy": 0.200591941177845,
      "num_tokens": 39083284.0,
      "step": 22525
    },
    {
      "entropy": 5.666073226928711,
      "epoch": 1.7529274460221747,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004690872017720163,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19876710772514344,
      "num_tokens": 39091106.0,
      "step": 22530
    },
    {
      "entropy": 5.637571716308594,
      "epoch": 1.7533164753938921,
      "grad_norm": 1.125,
      "learning_rate": 0.00046907317981797006,
      "loss": 4.9933,
      "mean_token_accuracy": 0.20645665377378464,
      "num_tokens": 39100499.0,
      "step": 22535
    },
    {
      "entropy": 5.65658974647522,
      "epoch": 1.7537055047656098,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046905915491917213,
      "loss": 5.1054,
      "mean_token_accuracy": 0.20423993915319444,
      "num_tokens": 39108895.0,
      "step": 22540
    },
    {
      "entropy": 5.537496185302734,
      "epoch": 1.7540945341373275,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004690451270758354,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21444785743951797,
      "num_tokens": 39117053.0,
      "step": 22545
    },
    {
      "entropy": 5.620280408859253,
      "epoch": 1.7544835635090448,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046903109628817276,
      "loss": 5.0603,
      "mean_token_accuracy": 0.2013158842921257,
      "num_tokens": 39125750.0,
      "step": 22550
    },
    {
      "entropy": 5.649320983886719,
      "epoch": 1.7548725928807625,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046901706255639703,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2014069750905037,
      "num_tokens": 39134105.0,
      "step": 22555
    },
    {
      "entropy": 5.744879627227784,
      "epoch": 1.7552616222524802,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004690030258807212,
      "loss": 5.201,
      "mean_token_accuracy": 0.19582940340042115,
      "num_tokens": 39143543.0,
      "step": 22560
    },
    {
      "entropy": 5.671559429168701,
      "epoch": 1.7556506516241976,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046898898626135827,
      "loss": 5.0922,
      "mean_token_accuracy": 0.19440879076719284,
      "num_tokens": 39152826.0,
      "step": 22565
    },
    {
      "entropy": 5.677198791503907,
      "epoch": 1.7560396809959151,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004689749436985211,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19286631047725677,
      "num_tokens": 39160999.0,
      "step": 22570
    },
    {
      "entropy": 5.657620477676391,
      "epoch": 1.7564287103676328,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046896089819242304,
      "loss": 5.0569,
      "mean_token_accuracy": 0.1979947492480278,
      "num_tokens": 39169134.0,
      "step": 22575
    },
    {
      "entropy": 5.630805492401123,
      "epoch": 1.7568177397393503,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004689468497432771,
      "loss": 4.994,
      "mean_token_accuracy": 0.20295559912919997,
      "num_tokens": 39177545.0,
      "step": 22580
    },
    {
      "entropy": 5.588210344314575,
      "epoch": 1.7572067691110678,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004689327983512963,
      "loss": 4.971,
      "mean_token_accuracy": 0.20277802050113677,
      "num_tokens": 39185671.0,
      "step": 22585
    },
    {
      "entropy": 5.645377731323242,
      "epoch": 1.7575957984827855,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046891874401669404,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1919768735766411,
      "num_tokens": 39194420.0,
      "step": 22590
    },
    {
      "entropy": 5.636783409118652,
      "epoch": 1.7579848278545032,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004689046867396834,
      "loss": 5.0571,
      "mean_token_accuracy": 0.1989524707198143,
      "num_tokens": 39203295.0,
      "step": 22595
    },
    {
      "entropy": 5.605256509780884,
      "epoch": 1.7583738572262204,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004688906265204779,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2007482960820198,
      "num_tokens": 39211838.0,
      "step": 22600
    },
    {
      "entropy": 5.652195644378662,
      "epoch": 1.7587628865979381,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004688765633592907,
      "loss": 4.9704,
      "mean_token_accuracy": 0.19742825478315354,
      "num_tokens": 39219842.0,
      "step": 22605
    },
    {
      "entropy": 5.628628158569336,
      "epoch": 1.7591519159696558,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004688624972563352,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20181595385074616,
      "num_tokens": 39228167.0,
      "step": 22610
    },
    {
      "entropy": 5.620701646804809,
      "epoch": 1.7595409453413733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004688484282118249,
      "loss": 5.0953,
      "mean_token_accuracy": 0.19199856519699096,
      "num_tokens": 39236546.0,
      "step": 22615
    },
    {
      "entropy": 5.652436113357544,
      "epoch": 1.7599299747130908,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004688343562259732,
      "loss": 5.0464,
      "mean_token_accuracy": 0.20907357931137086,
      "num_tokens": 39244212.0,
      "step": 22620
    },
    {
      "entropy": 5.6109332084655765,
      "epoch": 1.7603190040848085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004688202812989937,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20230043530464173,
      "num_tokens": 39252892.0,
      "step": 22625
    },
    {
      "entropy": 5.5753981590271,
      "epoch": 1.760708033456526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004688062034311,
      "loss": 4.9917,
      "mean_token_accuracy": 0.20682133436203004,
      "num_tokens": 39262261.0,
      "step": 22630
    },
    {
      "entropy": 5.693252754211426,
      "epoch": 1.7610970628282434,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046879212262250546,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20017526298761368,
      "num_tokens": 39270790.0,
      "step": 22635
    },
    {
      "entropy": 5.618958330154419,
      "epoch": 1.761486092199961,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046877803887342406,
      "loss": 5.0495,
      "mean_token_accuracy": 0.19491275995969773,
      "num_tokens": 39278825.0,
      "step": 22640
    },
    {
      "entropy": 5.6251115798950195,
      "epoch": 1.7618751215716788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004687639521840693,
      "loss": 5.0544,
      "mean_token_accuracy": 0.19777677059173585,
      "num_tokens": 39288343.0,
      "step": 22645
    },
    {
      "entropy": 5.705674123764038,
      "epoch": 1.7622641509433963,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046874986255465495,
      "loss": 5.1065,
      "mean_token_accuracy": 0.19336846172809602,
      "num_tokens": 39296562.0,
      "step": 22650
    },
    {
      "entropy": 5.66304759979248,
      "epoch": 1.7626531803151138,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046873576998539485,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20383942425251006,
      "num_tokens": 39305433.0,
      "step": 22655
    },
    {
      "entropy": 5.62990255355835,
      "epoch": 1.7630422096868315,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004687216744765028,
      "loss": 5.1073,
      "mean_token_accuracy": 0.1918163388967514,
      "num_tokens": 39313244.0,
      "step": 22660
    },
    {
      "entropy": 5.687591934204102,
      "epoch": 1.763431239058549,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004687075760281927,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19727670401334763,
      "num_tokens": 39321933.0,
      "step": 22665
    },
    {
      "entropy": 5.702593898773193,
      "epoch": 1.7638202684302664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004686934746406784,
      "loss": 5.0905,
      "mean_token_accuracy": 0.2011447563767433,
      "num_tokens": 39331194.0,
      "step": 22670
    },
    {
      "entropy": 5.687296676635742,
      "epoch": 1.764209297801984,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046867937031417397,
      "loss": 5.018,
      "mean_token_accuracy": 0.19477686583995818,
      "num_tokens": 39339762.0,
      "step": 22675
    },
    {
      "entropy": 5.642330884933472,
      "epoch": 1.7645983271737016,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004686652630488933,
      "loss": 5.0856,
      "mean_token_accuracy": 0.20060114115476607,
      "num_tokens": 39349020.0,
      "step": 22680
    },
    {
      "entropy": 5.71389479637146,
      "epoch": 1.764987356545419,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046865115284505063,
      "loss": 5.2119,
      "mean_token_accuracy": 0.18933181017637252,
      "num_tokens": 39357889.0,
      "step": 22685
    },
    {
      "entropy": 5.666406059265137,
      "epoch": 1.7653763859171367,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046863703970285986,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20060369819402696,
      "num_tokens": 39366120.0,
      "step": 22690
    },
    {
      "entropy": 5.5963959217071535,
      "epoch": 1.7657654152888544,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004686229236225352,
      "loss": 4.9091,
      "mean_token_accuracy": 0.20524757355451584,
      "num_tokens": 39374056.0,
      "step": 22695
    },
    {
      "entropy": 5.654457855224609,
      "epoch": 1.766154444660572,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004686088046042909,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19870393425226213,
      "num_tokens": 39383170.0,
      "step": 22700
    },
    {
      "entropy": 5.6901051044464115,
      "epoch": 1.7665434740322894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046859468264834114,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19672829806804656,
      "num_tokens": 39391599.0,
      "step": 22705
    },
    {
      "entropy": 5.723700428009034,
      "epoch": 1.766932503404007,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046858055775490017,
      "loss": 5.1325,
      "mean_token_accuracy": 0.19089903235435485,
      "num_tokens": 39400217.0,
      "step": 22710
    },
    {
      "entropy": 5.684020280838013,
      "epoch": 1.7673215327757246,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004685664299241825,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19731210619211198,
      "num_tokens": 39408943.0,
      "step": 22715
    },
    {
      "entropy": 5.664070749282837,
      "epoch": 1.767710562147442,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004685522991564022,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19333985447883606,
      "num_tokens": 39417361.0,
      "step": 22720
    },
    {
      "entropy": 5.4988175392150875,
      "epoch": 1.7680995915191597,
      "grad_norm": 1.125,
      "learning_rate": 0.0004685381654517738,
      "loss": 4.8358,
      "mean_token_accuracy": 0.20954587906599045,
      "num_tokens": 39425896.0,
      "step": 22725
    },
    {
      "entropy": 5.6054259777069095,
      "epoch": 1.7684886208908772,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004685240288105119,
      "loss": 5.05,
      "mean_token_accuracy": 0.20338875502347947,
      "num_tokens": 39434719.0,
      "step": 22730
    },
    {
      "entropy": 5.705852746963501,
      "epoch": 1.7688776502625947,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046850988923283085,
      "loss": 5.0093,
      "mean_token_accuracy": 0.19465023577213286,
      "num_tokens": 39443731.0,
      "step": 22735
    },
    {
      "entropy": 5.722311735153198,
      "epoch": 1.7692666796343124,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046849574671894523,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19653709828853608,
      "num_tokens": 39452430.0,
      "step": 22740
    },
    {
      "entropy": 5.669995498657227,
      "epoch": 1.76965570900603,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046848160126906956,
      "loss": 4.9617,
      "mean_token_accuracy": 0.20337569415569307,
      "num_tokens": 39460560.0,
      "step": 22745
    },
    {
      "entropy": 5.677088975906372,
      "epoch": 1.7700447383777476,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004684674528834186,
      "loss": 5.0857,
      "mean_token_accuracy": 0.19205645620822906,
      "num_tokens": 39469575.0,
      "step": 22750
    },
    {
      "entropy": 5.633237266540528,
      "epoch": 1.770433767749465,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046845330156220703,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2004252478480339,
      "num_tokens": 39478297.0,
      "step": 22755
    },
    {
      "entropy": 5.679377794265747,
      "epoch": 1.7708227971211827,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004684391473056495,
      "loss": 5.0925,
      "mean_token_accuracy": 0.1932470127940178,
      "num_tokens": 39486947.0,
      "step": 22760
    },
    {
      "entropy": 5.672457695007324,
      "epoch": 1.7712118264929002,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046842499011396076,
      "loss": 5.0713,
      "mean_token_accuracy": 0.19731754958629608,
      "num_tokens": 39495649.0,
      "step": 22765
    },
    {
      "entropy": 5.641335535049438,
      "epoch": 1.7716008558646177,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046841082998735575,
      "loss": 5.0545,
      "mean_token_accuracy": 0.19844259172677994,
      "num_tokens": 39504733.0,
      "step": 22770
    },
    {
      "entropy": 5.579748725891113,
      "epoch": 1.7719898852363354,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046839666692604916,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20175923109054567,
      "num_tokens": 39513007.0,
      "step": 22775
    },
    {
      "entropy": 5.668133211135864,
      "epoch": 1.7723789146080529,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000468382500930256,
      "loss": 5.1322,
      "mean_token_accuracy": 0.2032559707760811,
      "num_tokens": 39521734.0,
      "step": 22780
    },
    {
      "entropy": 5.728643417358398,
      "epoch": 1.7727679439797703,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046836833200019116,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19597928822040558,
      "num_tokens": 39530525.0,
      "step": 22785
    },
    {
      "entropy": 5.697123670578003,
      "epoch": 1.773156973351488,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004683541601360697,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20339512526988984,
      "num_tokens": 39538938.0,
      "step": 22790
    },
    {
      "entropy": 5.566825914382934,
      "epoch": 1.7735460027232057,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046833998533810653,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21193620562553406,
      "num_tokens": 39547342.0,
      "step": 22795
    },
    {
      "entropy": 5.593465423583984,
      "epoch": 1.7739350320949232,
      "grad_norm": 1.25,
      "learning_rate": 0.0004683258076065169,
      "loss": 5.027,
      "mean_token_accuracy": 0.19637676030397416,
      "num_tokens": 39555107.0,
      "step": 22800
    },
    {
      "entropy": 5.6839131832122805,
      "epoch": 1.7743240614666407,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046831162694151586,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19328555911779405,
      "num_tokens": 39563226.0,
      "step": 22805
    },
    {
      "entropy": 5.696634769439697,
      "epoch": 1.7747130908383584,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046829744334331855,
      "loss": 5.0964,
      "mean_token_accuracy": 0.19820945858955383,
      "num_tokens": 39571689.0,
      "step": 22810
    },
    {
      "entropy": 5.6443298816680905,
      "epoch": 1.7751021202100759,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046828325681214013,
      "loss": 5.0137,
      "mean_token_accuracy": 0.20227995365858079,
      "num_tokens": 39579985.0,
      "step": 22815
    },
    {
      "entropy": 5.682964944839478,
      "epoch": 1.7754911495817933,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000468269067348196,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1938783898949623,
      "num_tokens": 39587892.0,
      "step": 22820
    },
    {
      "entropy": 5.653334903717041,
      "epoch": 1.775880178953511,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004682548749517014,
      "loss": 5.1386,
      "mean_token_accuracy": 0.19406886845827104,
      "num_tokens": 39597516.0,
      "step": 22825
    },
    {
      "entropy": 5.706963729858399,
      "epoch": 1.7762692083252287,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046824067962287163,
      "loss": 5.0678,
      "mean_token_accuracy": 0.19957088232040404,
      "num_tokens": 39605860.0,
      "step": 22830
    },
    {
      "entropy": 5.69265775680542,
      "epoch": 1.776658237696946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004682264813619221,
      "loss": 5.111,
      "mean_token_accuracy": 0.19365580081939698,
      "num_tokens": 39614915.0,
      "step": 22835
    },
    {
      "entropy": 5.7231574058532715,
      "epoch": 1.7770472670686637,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046821228016906836,
      "loss": 5.0828,
      "mean_token_accuracy": 0.20129266381263733,
      "num_tokens": 39623588.0,
      "step": 22840
    },
    {
      "entropy": 5.70831561088562,
      "epoch": 1.7774362964403814,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004681980760445257,
      "loss": 5.1247,
      "mean_token_accuracy": 0.19598506689071654,
      "num_tokens": 39632351.0,
      "step": 22845
    },
    {
      "entropy": 5.633097171783447,
      "epoch": 1.7778253258120988,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004681838689885098,
      "loss": 5.06,
      "mean_token_accuracy": 0.1998940736055374,
      "num_tokens": 39641053.0,
      "step": 22850
    },
    {
      "entropy": 5.572714757919312,
      "epoch": 1.7782143551838163,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004681696590012362,
      "loss": 4.906,
      "mean_token_accuracy": 0.20548200011253356,
      "num_tokens": 39649838.0,
      "step": 22855
    },
    {
      "entropy": 5.607601928710937,
      "epoch": 1.778603384555534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046815544608292043,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2000594988465309,
      "num_tokens": 39658886.0,
      "step": 22860
    },
    {
      "entropy": 5.696227645874023,
      "epoch": 1.7789924139272515,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004681412302337782,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20609269589185714,
      "num_tokens": 39667279.0,
      "step": 22865
    },
    {
      "entropy": 5.635987997055054,
      "epoch": 1.779381443298969,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004681270114540252,
      "loss": 5.0291,
      "mean_token_accuracy": 0.19643961191177367,
      "num_tokens": 39676213.0,
      "step": 22870
    },
    {
      "entropy": 5.653455305099487,
      "epoch": 1.7797704726706867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004681127897438772,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20093480497598648,
      "num_tokens": 39685859.0,
      "step": 22875
    },
    {
      "entropy": 5.602489042282104,
      "epoch": 1.7801595020424044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046809856510355007,
      "loss": 4.9861,
      "mean_token_accuracy": 0.19861941635608674,
      "num_tokens": 39695240.0,
      "step": 22880
    },
    {
      "entropy": 5.645137786865234,
      "epoch": 1.7805485314141216,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004680843375332595,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2028212308883667,
      "num_tokens": 39704441.0,
      "step": 22885
    },
    {
      "entropy": 5.7116001605987545,
      "epoch": 1.7809375607858393,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046807010703322143,
      "loss": 5.1056,
      "mean_token_accuracy": 0.19841129332780838,
      "num_tokens": 39714074.0,
      "step": 22890
    },
    {
      "entropy": 5.741708993911743,
      "epoch": 1.781326590157557,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004680558736036518,
      "loss": 5.105,
      "mean_token_accuracy": 0.1881596788764,
      "num_tokens": 39722527.0,
      "step": 22895
    },
    {
      "entropy": 5.675631809234619,
      "epoch": 1.7817156195292745,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004680416372447666,
      "loss": 5.0789,
      "mean_token_accuracy": 0.19760531783103943,
      "num_tokens": 39730983.0,
      "step": 22900
    },
    {
      "entropy": 5.717896175384522,
      "epoch": 1.782104648900992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046802739795678177,
      "loss": 5.14,
      "mean_token_accuracy": 0.1933130294084549,
      "num_tokens": 39741114.0,
      "step": 22905
    },
    {
      "entropy": 5.716539144515991,
      "epoch": 1.7824936782727097,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046801315573991346,
      "loss": 5.0483,
      "mean_token_accuracy": 0.1988396853208542,
      "num_tokens": 39749281.0,
      "step": 22910
    },
    {
      "entropy": 5.638431358337402,
      "epoch": 1.7828827076444271,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046799891059437764,
      "loss": 4.9163,
      "mean_token_accuracy": 0.20659107118844985,
      "num_tokens": 39758136.0,
      "step": 22915
    },
    {
      "entropy": 5.6456811904907225,
      "epoch": 1.7832717370161446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046798466252039056,
      "loss": 5.1032,
      "mean_token_accuracy": 0.20015140622854233,
      "num_tokens": 39767104.0,
      "step": 22920
    },
    {
      "entropy": 5.655909442901612,
      "epoch": 1.7836607663878623,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046797041151816845,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20372353941202165,
      "num_tokens": 39775979.0,
      "step": 22925
    },
    {
      "entropy": 5.6257688999176025,
      "epoch": 1.78404979575958,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046795615758792747,
      "loss": 4.9335,
      "mean_token_accuracy": 0.20386708676815032,
      "num_tokens": 39783740.0,
      "step": 22930
    },
    {
      "entropy": 5.634879302978516,
      "epoch": 1.7844388251312973,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004679419007298839,
      "loss": 5.0764,
      "mean_token_accuracy": 0.206395722925663,
      "num_tokens": 39791836.0,
      "step": 22935
    },
    {
      "entropy": 5.64626111984253,
      "epoch": 1.784827854503015,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004679276409442541,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20402800440788268,
      "num_tokens": 39800694.0,
      "step": 22940
    },
    {
      "entropy": 5.641003608703613,
      "epoch": 1.7852168838747327,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004679133782312544,
      "loss": 5.1069,
      "mean_token_accuracy": 0.19544054418802262,
      "num_tokens": 39809108.0,
      "step": 22945
    },
    {
      "entropy": 5.65320725440979,
      "epoch": 1.7856059132464501,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004678991125911013,
      "loss": 5.0175,
      "mean_token_accuracy": 0.19981946796178818,
      "num_tokens": 39817608.0,
      "step": 22950
    },
    {
      "entropy": 5.654651212692261,
      "epoch": 1.7859949426181676,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004678848440240111,
      "loss": 5.006,
      "mean_token_accuracy": 0.20390069037675856,
      "num_tokens": 39826596.0,
      "step": 22955
    },
    {
      "entropy": 5.6365326881408695,
      "epoch": 1.7863839719898853,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004678705725302005,
      "loss": 5.0235,
      "mean_token_accuracy": 0.1989862784743309,
      "num_tokens": 39834440.0,
      "step": 22960
    },
    {
      "entropy": 5.6035261154174805,
      "epoch": 1.7867730013616028,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004678562981098859,
      "loss": 5.0713,
      "mean_token_accuracy": 0.19499050676822663,
      "num_tokens": 39842617.0,
      "step": 22965
    },
    {
      "entropy": 5.573136854171753,
      "epoch": 1.7871620307333203,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046784202076328394,
      "loss": 4.942,
      "mean_token_accuracy": 0.21251089721918107,
      "num_tokens": 39850688.0,
      "step": 22970
    },
    {
      "entropy": 5.658451271057129,
      "epoch": 1.787551060105038,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046782774049061124,
      "loss": 5.046,
      "mean_token_accuracy": 0.19893334358930587,
      "num_tokens": 39859347.0,
      "step": 22975
    },
    {
      "entropy": 5.623265361785888,
      "epoch": 1.7879400894767556,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004678134572920845,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20410974323749542,
      "num_tokens": 39867090.0,
      "step": 22980
    },
    {
      "entropy": 5.6374561309814455,
      "epoch": 1.788329118848473,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004677991711679204,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20177548825740815,
      "num_tokens": 39875653.0,
      "step": 22985
    },
    {
      "entropy": 5.6455625057220455,
      "epoch": 1.7887181482201906,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046778488211833585,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19069569706916809,
      "num_tokens": 39884781.0,
      "step": 22990
    },
    {
      "entropy": 5.675390434265137,
      "epoch": 1.7891071775919083,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004677705901435475,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20029411017894744,
      "num_tokens": 39892619.0,
      "step": 22995
    },
    {
      "entropy": 5.658169317245483,
      "epoch": 1.7894962069636258,
      "grad_norm": 1.125,
      "learning_rate": 0.0004677562952437723,
      "loss": 5.0496,
      "mean_token_accuracy": 0.19682631492614747,
      "num_tokens": 39901518.0,
      "step": 23000
    },
    {
      "entropy": 5.627274513244629,
      "epoch": 1.7898852363353432,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046774199741922705,
      "loss": 5.0099,
      "mean_token_accuracy": 0.2083907753229141,
      "num_tokens": 39909550.0,
      "step": 23005
    },
    {
      "entropy": 5.680173635482788,
      "epoch": 1.790274265707061,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046772769667012884,
      "loss": 5.0619,
      "mean_token_accuracy": 0.2035966172814369,
      "num_tokens": 39918177.0,
      "step": 23010
    },
    {
      "entropy": 5.628311681747436,
      "epoch": 1.7906632950787784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004677133929966946,
      "loss": 5.0499,
      "mean_token_accuracy": 0.19588937163352965,
      "num_tokens": 39926812.0,
      "step": 23015
    },
    {
      "entropy": 5.647428226470947,
      "epoch": 1.791052324450496,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046769908639914134,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1968407303094864,
      "num_tokens": 39934506.0,
      "step": 23020
    },
    {
      "entropy": 5.591739130020142,
      "epoch": 1.7914413538222136,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004676847768776861,
      "loss": 5.0554,
      "mean_token_accuracy": 0.19508164823055268,
      "num_tokens": 39943252.0,
      "step": 23025
    },
    {
      "entropy": 5.643436622619629,
      "epoch": 1.7918303831939313,
      "grad_norm": 1.25,
      "learning_rate": 0.0004676704644325462,
      "loss": 5.0675,
      "mean_token_accuracy": 0.19665274024009705,
      "num_tokens": 39952585.0,
      "step": 23030
    },
    {
      "entropy": 5.703285264968872,
      "epoch": 1.7922194125656488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004676561490639385,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20355939269065856,
      "num_tokens": 39961805.0,
      "step": 23035
    },
    {
      "entropy": 5.688137197494507,
      "epoch": 1.7926084419373662,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004676418307720805,
      "loss": 5.0672,
      "mean_token_accuracy": 0.2016652137041092,
      "num_tokens": 39970045.0,
      "step": 23040
    },
    {
      "entropy": 5.679377841949463,
      "epoch": 1.792997471309084,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004676275095571892,
      "loss": 5.0167,
      "mean_token_accuracy": 0.1981065183877945,
      "num_tokens": 39979073.0,
      "step": 23045
    },
    {
      "entropy": 5.668692350387573,
      "epoch": 1.7933865006808014,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046761318541948215,
      "loss": 5.0372,
      "mean_token_accuracy": 0.20204666554927825,
      "num_tokens": 39987936.0,
      "step": 23050
    },
    {
      "entropy": 5.587795066833496,
      "epoch": 1.7937755300525189,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004675988583591766,
      "loss": 5.0728,
      "mean_token_accuracy": 0.19786328971385955,
      "num_tokens": 39996547.0,
      "step": 23055
    },
    {
      "entropy": 5.6756048679351805,
      "epoch": 1.7941645594242366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046758452837648997,
      "loss": 5.1288,
      "mean_token_accuracy": 0.18661812096834182,
      "num_tokens": 40005717.0,
      "step": 23060
    },
    {
      "entropy": 5.65836353302002,
      "epoch": 1.794553588795954,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004675701954716395,
      "loss": 4.9395,
      "mean_token_accuracy": 0.20196045637130738,
      "num_tokens": 40014799.0,
      "step": 23065
    },
    {
      "entropy": 5.678746080398559,
      "epoch": 1.7949426181676715,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046755585964484286,
      "loss": 5.0298,
      "mean_token_accuracy": 0.1966300755739212,
      "num_tokens": 40023899.0,
      "step": 23070
    },
    {
      "entropy": 5.6207013607025145,
      "epoch": 1.7953316475393892,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004675415208963177,
      "loss": 5.0096,
      "mean_token_accuracy": 0.203628271818161,
      "num_tokens": 40032629.0,
      "step": 23075
    },
    {
      "entropy": 5.658460855484009,
      "epoch": 1.795720676911107,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046752717922628125,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20494168251752853,
      "num_tokens": 40040575.0,
      "step": 23080
    },
    {
      "entropy": 5.664865922927857,
      "epoch": 1.7961097062828244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004675128346349514,
      "loss": 5.0915,
      "mean_token_accuracy": 0.19283560514450074,
      "num_tokens": 40049221.0,
      "step": 23085
    },
    {
      "entropy": 5.69799771308899,
      "epoch": 1.7964987356545419,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046749848712254554,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1930222898721695,
      "num_tokens": 40057737.0,
      "step": 23090
    },
    {
      "entropy": 5.618962383270263,
      "epoch": 1.7968877650262596,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046748413668928164,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20602759718894958,
      "num_tokens": 40066447.0,
      "step": 23095
    },
    {
      "entropy": 5.6520774364471436,
      "epoch": 1.797276794397977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004674697833353774,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19726436883211135,
      "num_tokens": 40075825.0,
      "step": 23100
    },
    {
      "entropy": 5.683380937576294,
      "epoch": 1.7976658237696945,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046745542706105045,
      "loss": 5.101,
      "mean_token_accuracy": 0.19633927792310715,
      "num_tokens": 40084342.0,
      "step": 23105
    },
    {
      "entropy": 5.747971248626709,
      "epoch": 1.7980548531414122,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046744106786651875,
      "loss": 5.1065,
      "mean_token_accuracy": 0.1895572602748871,
      "num_tokens": 40092650.0,
      "step": 23110
    },
    {
      "entropy": 5.623660230636597,
      "epoch": 1.7984438825131297,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004674267057520001,
      "loss": 5.043,
      "mean_token_accuracy": 0.19365017414093016,
      "num_tokens": 40100738.0,
      "step": 23115
    },
    {
      "entropy": 5.5566259860992435,
      "epoch": 1.7988329118848472,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004674123407177125,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20417187511920928,
      "num_tokens": 40109213.0,
      "step": 23120
    },
    {
      "entropy": 5.608888339996338,
      "epoch": 1.7992219412565649,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046739797276387394,
      "loss": 4.889,
      "mean_token_accuracy": 0.21043968051671982,
      "num_tokens": 40117622.0,
      "step": 23125
    },
    {
      "entropy": 5.64186110496521,
      "epoch": 1.7996109706282826,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046738360189070235,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19907923489809037,
      "num_tokens": 40126347.0,
      "step": 23130
    },
    {
      "entropy": 5.757534790039062,
      "epoch": 1.8,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004673692280984157,
      "loss": 5.1445,
      "mean_token_accuracy": 0.18662794679403305,
      "num_tokens": 40134245.0,
      "step": 23135
    },
    {
      "entropy": 5.667551183700562,
      "epoch": 1.8003890293717175,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004673548513872324,
      "loss": 4.9898,
      "mean_token_accuracy": 0.1926770031452179,
      "num_tokens": 40142304.0,
      "step": 23140
    },
    {
      "entropy": 5.635939216613769,
      "epoch": 1.8007780587434352,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046734047175737026,
      "loss": 4.9987,
      "mean_token_accuracy": 0.19706318974494935,
      "num_tokens": 40150999.0,
      "step": 23145
    },
    {
      "entropy": 5.68206901550293,
      "epoch": 1.8011670881151527,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046732608920904754,
      "loss": 5.033,
      "mean_token_accuracy": 0.19831780940294266,
      "num_tokens": 40158659.0,
      "step": 23150
    },
    {
      "entropy": 5.5862266540527346,
      "epoch": 1.8015561174868702,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004673117037424827,
      "loss": 4.9561,
      "mean_token_accuracy": 0.20681921988725663,
      "num_tokens": 40167343.0,
      "step": 23155
    },
    {
      "entropy": 5.637562656402588,
      "epoch": 1.8019451468585879,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046729731535789375,
      "loss": 5.0224,
      "mean_token_accuracy": 0.20230080038309098,
      "num_tokens": 40176433.0,
      "step": 23160
    },
    {
      "entropy": 5.725575590133667,
      "epoch": 1.8023341762303053,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046728292405549913,
      "loss": 5.0842,
      "mean_token_accuracy": 0.19801634401082993,
      "num_tokens": 40185826.0,
      "step": 23165
    },
    {
      "entropy": 5.655608177185059,
      "epoch": 1.8027232056020228,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004672685298355172,
      "loss": 5.1039,
      "mean_token_accuracy": 0.19996628016233445,
      "num_tokens": 40194359.0,
      "step": 23170
    },
    {
      "entropy": 5.592953777313232,
      "epoch": 1.8031122349737405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004672541326981664,
      "loss": 4.9192,
      "mean_token_accuracy": 0.20581860989332199,
      "num_tokens": 40202812.0,
      "step": 23175
    },
    {
      "entropy": 5.653076648712158,
      "epoch": 1.8035012643454582,
      "grad_norm": 1.109375,
      "learning_rate": 0.000467239732643665,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19182265549898148,
      "num_tokens": 40212344.0,
      "step": 23180
    },
    {
      "entropy": 5.685462474822998,
      "epoch": 1.8038902937171757,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046722532967223183,
      "loss": 5.1388,
      "mean_token_accuracy": 0.19260006994009018,
      "num_tokens": 40220974.0,
      "step": 23185
    },
    {
      "entropy": 5.6697290420532225,
      "epoch": 1.8042793230888932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004672109237840851,
      "loss": 5.0828,
      "mean_token_accuracy": 0.19274238795042037,
      "num_tokens": 40229616.0,
      "step": 23190
    },
    {
      "entropy": 5.679400634765625,
      "epoch": 1.8046683524606109,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046719651497944355,
      "loss": 5.0692,
      "mean_token_accuracy": 0.19668403416872024,
      "num_tokens": 40238505.0,
      "step": 23195
    },
    {
      "entropy": 5.77176570892334,
      "epoch": 1.8050573818323283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004671821032585259,
      "loss": 5.192,
      "mean_token_accuracy": 0.19094980806112288,
      "num_tokens": 40246716.0,
      "step": 23200
    },
    {
      "entropy": 5.706794500350952,
      "epoch": 1.8054464112040458,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004671676886215506,
      "loss": 5.1511,
      "mean_token_accuracy": 0.18593613654375077,
      "num_tokens": 40256146.0,
      "step": 23205
    },
    {
      "entropy": 5.7128541469573975,
      "epoch": 1.8058354405757635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004671532710687365,
      "loss": 5.0719,
      "mean_token_accuracy": 0.20557624846696854,
      "num_tokens": 40265304.0,
      "step": 23210
    },
    {
      "entropy": 5.614098310470581,
      "epoch": 1.806224469947481,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004671388506003024,
      "loss": 4.9185,
      "mean_token_accuracy": 0.20143164098262786,
      "num_tokens": 40274055.0,
      "step": 23215
    },
    {
      "entropy": 5.653295278549194,
      "epoch": 1.8066134993191985,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004671244272164671,
      "loss": 4.99,
      "mean_token_accuracy": 0.1972822740674019,
      "num_tokens": 40283184.0,
      "step": 23220
    },
    {
      "entropy": 5.7943868160247805,
      "epoch": 1.8070025286909162,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046711000091744935,
      "loss": 5.1124,
      "mean_token_accuracy": 0.19664375185966493,
      "num_tokens": 40292604.0,
      "step": 23225
    },
    {
      "entropy": 5.582326173782349,
      "epoch": 1.8073915580626339,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004670955717034681,
      "loss": 4.8435,
      "mean_token_accuracy": 0.21837232410907745,
      "num_tokens": 40301096.0,
      "step": 23230
    },
    {
      "entropy": 5.700474500656128,
      "epoch": 1.8077805874343513,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046708113957474233,
      "loss": 5.2101,
      "mean_token_accuracy": 0.1904939040541649,
      "num_tokens": 40309976.0,
      "step": 23235
    },
    {
      "entropy": 5.782407569885254,
      "epoch": 1.8081696168060688,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000467066704531491,
      "loss": 5.1699,
      "mean_token_accuracy": 0.19290914237499238,
      "num_tokens": 40318594.0,
      "step": 23240
    },
    {
      "entropy": 5.687819528579712,
      "epoch": 1.8085586461777865,
      "grad_norm": 1.125,
      "learning_rate": 0.0004670522665739332,
      "loss": 5.0314,
      "mean_token_accuracy": 0.19965442419052123,
      "num_tokens": 40326480.0,
      "step": 23245
    },
    {
      "entropy": 5.635049676895141,
      "epoch": 1.808947675549504,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004670378257022878,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20088362246751784,
      "num_tokens": 40334646.0,
      "step": 23250
    },
    {
      "entropy": 5.658460330963135,
      "epoch": 1.8093367049212215,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046702338191677414,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19904337376356124,
      "num_tokens": 40344007.0,
      "step": 23255
    },
    {
      "entropy": 5.723978424072266,
      "epoch": 1.8097257342929391,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004670089352176113,
      "loss": 5.1044,
      "mean_token_accuracy": 0.19323526173830033,
      "num_tokens": 40352075.0,
      "step": 23260
    },
    {
      "entropy": 5.656817245483398,
      "epoch": 1.8101147636646568,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046699448560501847,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20091743767261505,
      "num_tokens": 40360725.0,
      "step": 23265
    },
    {
      "entropy": 5.620440912246704,
      "epoch": 1.810503793036374,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004669800330792149,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20299917012453078,
      "num_tokens": 40369022.0,
      "step": 23270
    },
    {
      "entropy": 5.665176725387573,
      "epoch": 1.8108928224080918,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046696557764041994,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20151742547750473,
      "num_tokens": 40377624.0,
      "step": 23275
    },
    {
      "entropy": 5.6575541496276855,
      "epoch": 1.8112818517798095,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004669511192888528,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19970284700393676,
      "num_tokens": 40385678.0,
      "step": 23280
    },
    {
      "entropy": 5.6143005847930905,
      "epoch": 1.811670881151527,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046693665802473294,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20033340603113176,
      "num_tokens": 40394565.0,
      "step": 23285
    },
    {
      "entropy": 5.710658216476441,
      "epoch": 1.8120599105232444,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046692219384827986,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19456615149974824,
      "num_tokens": 40402910.0,
      "step": 23290
    },
    {
      "entropy": 5.666740751266479,
      "epoch": 1.8124489398949621,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004669077267597129,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2055998757481575,
      "num_tokens": 40411352.0,
      "step": 23295
    },
    {
      "entropy": 5.629631757736206,
      "epoch": 1.8128379692666796,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046689325675925174,
      "loss": 5.0936,
      "mean_token_accuracy": 0.2014686182141304,
      "num_tokens": 40419925.0,
      "step": 23300
    },
    {
      "entropy": 5.64031023979187,
      "epoch": 1.813226998638397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046687878384711574,
      "loss": 5.0634,
      "mean_token_accuracy": 0.19735018908977509,
      "num_tokens": 40428948.0,
      "step": 23305
    },
    {
      "entropy": 5.682642889022827,
      "epoch": 1.8136160280101148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046686430802352476,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1957295536994934,
      "num_tokens": 40437625.0,
      "step": 23310
    },
    {
      "entropy": 5.651401472091675,
      "epoch": 1.8140050573818325,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004668498292886982,
      "loss": 4.997,
      "mean_token_accuracy": 0.2024431511759758,
      "num_tokens": 40446435.0,
      "step": 23315
    },
    {
      "entropy": 5.7380307674407955,
      "epoch": 1.8143940867535497,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046683534764285577,
      "loss": 5.0797,
      "mean_token_accuracy": 0.19660846441984176,
      "num_tokens": 40455368.0,
      "step": 23320
    },
    {
      "entropy": 5.668961238861084,
      "epoch": 1.8147831161252674,
      "grad_norm": 1.125,
      "learning_rate": 0.00046682086308621725,
      "loss": 5.0629,
      "mean_token_accuracy": 0.19605904668569565,
      "num_tokens": 40464461.0,
      "step": 23325
    },
    {
      "entropy": 5.724513292312622,
      "epoch": 1.8151721454969851,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004668063756190026,
      "loss": 5.1403,
      "mean_token_accuracy": 0.18670904487371445,
      "num_tokens": 40473143.0,
      "step": 23330
    },
    {
      "entropy": 5.640878820419312,
      "epoch": 1.8155611748687026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046679188524143136,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20028386563062667,
      "num_tokens": 40481592.0,
      "step": 23335
    },
    {
      "entropy": 5.593812942504883,
      "epoch": 1.81595020424042,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004667773919537235,
      "loss": 4.9216,
      "mean_token_accuracy": 0.2078142985701561,
      "num_tokens": 40489445.0,
      "step": 23340
    },
    {
      "entropy": 5.655923700332641,
      "epoch": 1.8163392336121378,
      "grad_norm": 1.125,
      "learning_rate": 0.00046676289575609914,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20590965449810028,
      "num_tokens": 40498279.0,
      "step": 23345
    },
    {
      "entropy": 5.693883752822876,
      "epoch": 1.8167282629838553,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046674839664877796,
      "loss": 4.967,
      "mean_token_accuracy": 0.20663145035505295,
      "num_tokens": 40507051.0,
      "step": 23350
    },
    {
      "entropy": 5.6040033340454105,
      "epoch": 1.8171172923555727,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004667338946319801,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20682334452867507,
      "num_tokens": 40515213.0,
      "step": 23355
    },
    {
      "entropy": 5.640662622451782,
      "epoch": 1.8175063217272904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004667193897059255,
      "loss": 5.0779,
      "mean_token_accuracy": 0.19594551473855973,
      "num_tokens": 40523859.0,
      "step": 23360
    },
    {
      "entropy": 5.726602363586426,
      "epoch": 1.8178953510990081,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046670488187083436,
      "loss": 5.1508,
      "mean_token_accuracy": 0.19366347193717956,
      "num_tokens": 40533236.0,
      "step": 23365
    },
    {
      "entropy": 5.662904119491577,
      "epoch": 1.8182843804707254,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004666903711269267,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2015302836894989,
      "num_tokens": 40541670.0,
      "step": 23370
    },
    {
      "entropy": 5.654554557800293,
      "epoch": 1.818673409842443,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004666758574744228,
      "loss": 5.0744,
      "mean_token_accuracy": 0.2013315498828888,
      "num_tokens": 40551411.0,
      "step": 23375
    },
    {
      "entropy": 5.694829320907592,
      "epoch": 1.8190624392141608,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004666613409135429,
      "loss": 5.0482,
      "mean_token_accuracy": 0.19648958146572112,
      "num_tokens": 40560351.0,
      "step": 23380
    },
    {
      "entropy": 5.6153693199157715,
      "epoch": 1.8194514685858783,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004666468214445072,
      "loss": 4.9806,
      "mean_token_accuracy": 0.21134056597948075,
      "num_tokens": 40569363.0,
      "step": 23385
    },
    {
      "entropy": 5.6943529605865475,
      "epoch": 1.8198404979575957,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046663229906753595,
      "loss": 5.0528,
      "mean_token_accuracy": 0.1931375190615654,
      "num_tokens": 40577425.0,
      "step": 23390
    },
    {
      "entropy": 5.703410816192627,
      "epoch": 1.8202295273293134,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046661777378284965,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1931896388530731,
      "num_tokens": 40585718.0,
      "step": 23395
    },
    {
      "entropy": 5.777365446090698,
      "epoch": 1.820618556701031,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004666032455906685,
      "loss": 5.1488,
      "mean_token_accuracy": 0.19303836673498154,
      "num_tokens": 40594886.0,
      "step": 23400
    },
    {
      "entropy": 5.7587135314941404,
      "epoch": 1.8210075860727484,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046658871449121325,
      "loss": 5.1476,
      "mean_token_accuracy": 0.18933731764554979,
      "num_tokens": 40603737.0,
      "step": 23405
    },
    {
      "entropy": 5.679676294326782,
      "epoch": 1.821396615444466,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004665741804847041,
      "loss": 5.0395,
      "mean_token_accuracy": 0.19731849581003189,
      "num_tokens": 40612473.0,
      "step": 23410
    },
    {
      "entropy": 5.645179653167725,
      "epoch": 1.8217856448161838,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046655964357136164,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20160272270441054,
      "num_tokens": 40620517.0,
      "step": 23415
    },
    {
      "entropy": 5.690299034118652,
      "epoch": 1.822174674187901,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046654510375140657,
      "loss": 5.086,
      "mean_token_accuracy": 0.18951601386070252,
      "num_tokens": 40629530.0,
      "step": 23420
    },
    {
      "entropy": 5.641601705551148,
      "epoch": 1.8225637035596187,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004665305610250594,
      "loss": 4.9912,
      "mean_token_accuracy": 0.205650395154953,
      "num_tokens": 40638616.0,
      "step": 23425
    },
    {
      "entropy": 5.706247997283936,
      "epoch": 1.8229527329313364,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004665160153925408,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1911957785487175,
      "num_tokens": 40646945.0,
      "step": 23430
    },
    {
      "entropy": 5.6208000659942625,
      "epoch": 1.823341762303054,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046650146685407154,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20164287835359573,
      "num_tokens": 40654845.0,
      "step": 23435
    },
    {
      "entropy": 5.5358171463012695,
      "epoch": 1.8237307916747714,
      "grad_norm": 1.125,
      "learning_rate": 0.00046648691540987226,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21397641003131868,
      "num_tokens": 40663776.0,
      "step": 23440
    },
    {
      "entropy": 5.675170803070069,
      "epoch": 1.824119821046489,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046647236106016387,
      "loss": 4.9715,
      "mean_token_accuracy": 0.20658940374851226,
      "num_tokens": 40672071.0,
      "step": 23445
    },
    {
      "entropy": 5.674432611465454,
      "epoch": 1.8245088504182065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004664578038051672,
      "loss": 5.1451,
      "mean_token_accuracy": 0.19230934232473373,
      "num_tokens": 40681264.0,
      "step": 23450
    },
    {
      "entropy": 5.629675006866455,
      "epoch": 1.824897879789924,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000466443243645103,
      "loss": 5.0446,
      "mean_token_accuracy": 0.19849296361207963,
      "num_tokens": 40690091.0,
      "step": 23455
    },
    {
      "entropy": 5.7384673118591305,
      "epoch": 1.8252869091616417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004664286805801923,
      "loss": 5.1422,
      "mean_token_accuracy": 0.193589249253273,
      "num_tokens": 40699623.0,
      "step": 23460
    },
    {
      "entropy": 5.652368688583374,
      "epoch": 1.8256759385333594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004664141146106562,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2016828775405884,
      "num_tokens": 40708288.0,
      "step": 23465
    },
    {
      "entropy": 5.7087485790252686,
      "epoch": 1.8260649679050769,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046639954573671547,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19920191168785095,
      "num_tokens": 40716357.0,
      "step": 23470
    },
    {
      "entropy": 5.7190111637115475,
      "epoch": 1.8264539972767944,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046638497395859127,
      "loss": 5.1107,
      "mean_token_accuracy": 0.19638197124004364,
      "num_tokens": 40724775.0,
      "step": 23475
    },
    {
      "entropy": 5.6122551441192625,
      "epoch": 1.826843026648512,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004663703992765047,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2047294408082962,
      "num_tokens": 40733642.0,
      "step": 23480
    },
    {
      "entropy": 5.626034164428711,
      "epoch": 1.8272320560202295,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046635582169067693,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20413835644721984,
      "num_tokens": 40742201.0,
      "step": 23485
    },
    {
      "entropy": 5.610148668289185,
      "epoch": 1.827621085391947,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046634124120132915,
      "loss": 4.988,
      "mean_token_accuracy": 0.2013638511300087,
      "num_tokens": 40750379.0,
      "step": 23490
    },
    {
      "entropy": 5.6681865692138675,
      "epoch": 1.8280101147636647,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004663266578086826,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19413399398326875,
      "num_tokens": 40759232.0,
      "step": 23495
    },
    {
      "entropy": 5.634824800491333,
      "epoch": 1.8283991441353822,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004663120715129585,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19612460285425187,
      "num_tokens": 40768247.0,
      "step": 23500
    },
    {
      "entropy": 5.708069658279419,
      "epoch": 1.8287881735070997,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004662974823143783,
      "loss": 5.1235,
      "mean_token_accuracy": 0.19401054382324218,
      "num_tokens": 40777085.0,
      "step": 23505
    },
    {
      "entropy": 5.7357430934906,
      "epoch": 1.8291772028788174,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004662828902131632,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19387224167585373,
      "num_tokens": 40785466.0,
      "step": 23510
    },
    {
      "entropy": 5.678268480300903,
      "epoch": 1.829566232250535,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004662682952095348,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2004189074039459,
      "num_tokens": 40794165.0,
      "step": 23515
    },
    {
      "entropy": 5.640733814239502,
      "epoch": 1.8299552616222525,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046625369730371436,
      "loss": 5.0252,
      "mean_token_accuracy": 0.19711432307958604,
      "num_tokens": 40802497.0,
      "step": 23520
    },
    {
      "entropy": 5.658389854431152,
      "epoch": 1.83034429099397,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004662390964959235,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20656996667385102,
      "num_tokens": 40811022.0,
      "step": 23525
    },
    {
      "entropy": 5.717386484146118,
      "epoch": 1.8307333203656877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046622449278638375,
      "loss": 5.078,
      "mean_token_accuracy": 0.1962095871567726,
      "num_tokens": 40820430.0,
      "step": 23530
    },
    {
      "entropy": 5.62439866065979,
      "epoch": 1.8311223497374052,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004662098861753167,
      "loss": 5.0836,
      "mean_token_accuracy": 0.1995575651526451,
      "num_tokens": 40829245.0,
      "step": 23535
    },
    {
      "entropy": 5.581241512298584,
      "epoch": 1.8315113791091227,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046619527666294394,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20278009176254272,
      "num_tokens": 40838576.0,
      "step": 23540
    },
    {
      "entropy": 5.729488658905029,
      "epoch": 1.8319004084808403,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004661806642494872,
      "loss": 5.124,
      "mean_token_accuracy": 0.1921101689338684,
      "num_tokens": 40847097.0,
      "step": 23545
    },
    {
      "entropy": 5.6860795497894285,
      "epoch": 1.8322894378525578,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004661660489351681,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2117006301879883,
      "num_tokens": 40855135.0,
      "step": 23550
    },
    {
      "entropy": 5.703047609329223,
      "epoch": 1.8326784672242753,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004661514307202086,
      "loss": 5.1163,
      "mean_token_accuracy": 0.19583360105752945,
      "num_tokens": 40864218.0,
      "step": 23555
    },
    {
      "entropy": 5.698640394210815,
      "epoch": 1.833067496595993,
      "grad_norm": 1.125,
      "learning_rate": 0.00046613680960483036,
      "loss": 5.0711,
      "mean_token_accuracy": 0.19870685786008835,
      "num_tokens": 40873147.0,
      "step": 23560
    },
    {
      "entropy": 5.696729612350464,
      "epoch": 1.8334565259677107,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004661221855892552,
      "loss": 5.1318,
      "mean_token_accuracy": 0.1981499046087265,
      "num_tokens": 40881340.0,
      "step": 23565
    },
    {
      "entropy": 5.632967472076416,
      "epoch": 1.8338455553394282,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046610755867370506,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20088434219360352,
      "num_tokens": 40889246.0,
      "step": 23570
    },
    {
      "entropy": 5.677598524093628,
      "epoch": 1.8342345847111456,
      "grad_norm": 1.1875,
      "learning_rate": 0.000466092928858402,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20690640360116958,
      "num_tokens": 40898285.0,
      "step": 23575
    },
    {
      "entropy": 5.635863637924194,
      "epoch": 1.8346236140828633,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046607829614356787,
      "loss": 5.0898,
      "mean_token_accuracy": 0.19565698504447937,
      "num_tokens": 40907012.0,
      "step": 23580
    },
    {
      "entropy": 5.662146759033203,
      "epoch": 1.8350126434545808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004660636605294247,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20094168931245804,
      "num_tokens": 40915857.0,
      "step": 23585
    },
    {
      "entropy": 5.719312953948974,
      "epoch": 1.8354016728262983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004660490220161946,
      "loss": 5.0151,
      "mean_token_accuracy": 0.19975469261407852,
      "num_tokens": 40924364.0,
      "step": 23590
    },
    {
      "entropy": 5.579410648345947,
      "epoch": 1.835790702198016,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046603438060409966,
      "loss": 5.0276,
      "mean_token_accuracy": 0.19955301582813262,
      "num_tokens": 40934092.0,
      "step": 23595
    },
    {
      "entropy": 5.624386596679687,
      "epoch": 1.8361797315697335,
      "grad_norm": 1.171875,
      "learning_rate": 0.000466019736293362,
      "loss": 4.9156,
      "mean_token_accuracy": 0.20587907284498214,
      "num_tokens": 40942066.0,
      "step": 23600
    },
    {
      "entropy": 5.650084733963013,
      "epoch": 1.836568760941451,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046600508908420396,
      "loss": 5.046,
      "mean_token_accuracy": 0.1995385393500328,
      "num_tokens": 40951445.0,
      "step": 23605
    },
    {
      "entropy": 5.612366437911987,
      "epoch": 1.8369577903131686,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046599043897684766,
      "loss": 5.0175,
      "mean_token_accuracy": 0.2062175080180168,
      "num_tokens": 40960111.0,
      "step": 23610
    },
    {
      "entropy": 5.666700983047486,
      "epoch": 1.8373468196848863,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004659757859715155,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2113405093550682,
      "num_tokens": 40968036.0,
      "step": 23615
    },
    {
      "entropy": 5.586283874511719,
      "epoch": 1.8377358490566038,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046596113006842975,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20348426699638367,
      "num_tokens": 40977032.0,
      "step": 23620
    },
    {
      "entropy": 5.633977937698364,
      "epoch": 1.8381248784283213,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004659464712678128,
      "loss": 5.0307,
      "mean_token_accuracy": 0.1976197123527527,
      "num_tokens": 40985523.0,
      "step": 23625
    },
    {
      "entropy": 5.64246826171875,
      "epoch": 1.838513907800039,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004659318095698871,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2081204429268837,
      "num_tokens": 40993617.0,
      "step": 23630
    },
    {
      "entropy": 5.583817815780639,
      "epoch": 1.8389029371717565,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000465917144974875,
      "loss": 5.0043,
      "mean_token_accuracy": 0.19729857742786408,
      "num_tokens": 41002424.0,
      "step": 23635
    },
    {
      "entropy": 5.675986051559448,
      "epoch": 1.839291966543474,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004659024774829992,
      "loss": 5.106,
      "mean_token_accuracy": 0.18890606015920638,
      "num_tokens": 41011150.0,
      "step": 23640
    },
    {
      "entropy": 5.690023946762085,
      "epoch": 1.8396809959151916,
      "grad_norm": 1.140625,
      "learning_rate": 0.000465887807094482,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20088308304548264,
      "num_tokens": 41020112.0,
      "step": 23645
    },
    {
      "entropy": 5.6951676368713375,
      "epoch": 1.8400700252869093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046587313380954635,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1960156813263893,
      "num_tokens": 41029470.0,
      "step": 23650
    },
    {
      "entropy": 5.685599994659424,
      "epoch": 1.8404590546586266,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046585845762841453,
      "loss": 5.0634,
      "mean_token_accuracy": 0.19622621238231658,
      "num_tokens": 41038622.0,
      "step": 23655
    },
    {
      "entropy": 5.7481622219085695,
      "epoch": 1.8408480840303443,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004658437785513095,
      "loss": 5.1902,
      "mean_token_accuracy": 0.19177833199501038,
      "num_tokens": 41047573.0,
      "step": 23660
    },
    {
      "entropy": 5.690817832946777,
      "epoch": 1.841237113402062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004658290965784539,
      "loss": 4.9918,
      "mean_token_accuracy": 0.20273317098617555,
      "num_tokens": 41057084.0,
      "step": 23665
    },
    {
      "entropy": 5.591980457305908,
      "epoch": 1.8416261427737795,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004658144117100704,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2043921485543251,
      "num_tokens": 41065297.0,
      "step": 23670
    },
    {
      "entropy": 5.6256719589233395,
      "epoch": 1.842015172145497,
      "grad_norm": 1.25,
      "learning_rate": 0.00046579972394638204,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20674618184566498,
      "num_tokens": 41073748.0,
      "step": 23675
    },
    {
      "entropy": 5.692818737030029,
      "epoch": 1.8424042015172146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046578503328761146,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2053775295615196,
      "num_tokens": 41082304.0,
      "step": 23680
    },
    {
      "entropy": 5.695648908615112,
      "epoch": 1.842793230888932,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046577033973398173,
      "loss": 5.1324,
      "mean_token_accuracy": 0.19709999412298201,
      "num_tokens": 41091130.0,
      "step": 23685
    },
    {
      "entropy": 5.66106481552124,
      "epoch": 1.8431822602606496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004657556432857157,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19572065770626068,
      "num_tokens": 41100413.0,
      "step": 23690
    },
    {
      "entropy": 5.542829751968384,
      "epoch": 1.8435712896323673,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004657409439430364,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21099061667919158,
      "num_tokens": 41108582.0,
      "step": 23695
    },
    {
      "entropy": 5.571869993209839,
      "epoch": 1.843960319004085,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004657262417061669,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20157552659511566,
      "num_tokens": 41117067.0,
      "step": 23700
    },
    {
      "entropy": 5.620513296127319,
      "epoch": 1.8443493483758022,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004657115365753302,
      "loss": 5.0008,
      "mean_token_accuracy": 0.19803959429264067,
      "num_tokens": 41125559.0,
      "step": 23705
    },
    {
      "entropy": 5.648423910140991,
      "epoch": 1.84473837774752,
      "grad_norm": 1.125,
      "learning_rate": 0.00046569682855074953,
      "loss": 5.0335,
      "mean_token_accuracy": 0.19989023804664613,
      "num_tokens": 41134503.0,
      "step": 23710
    },
    {
      "entropy": 5.711759948730469,
      "epoch": 1.8451274071192376,
      "grad_norm": 1.125,
      "learning_rate": 0.00046568211763264796,
      "loss": 5.1471,
      "mean_token_accuracy": 0.18766844272613525,
      "num_tokens": 41143903.0,
      "step": 23715
    },
    {
      "entropy": 5.633479452133178,
      "epoch": 1.845516436490955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004656674038212488,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19833906888961791,
      "num_tokens": 41153082.0,
      "step": 23720
    },
    {
      "entropy": 5.714179801940918,
      "epoch": 1.8459054658626726,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046565268711677525,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20792423486709594,
      "num_tokens": 41161749.0,
      "step": 23725
    },
    {
      "entropy": 5.6698071479797365,
      "epoch": 1.8462944952343903,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046563796751945065,
      "loss": 5.038,
      "mean_token_accuracy": 0.20310945063829422,
      "num_tokens": 41170969.0,
      "step": 23730
    },
    {
      "entropy": 5.576027727127075,
      "epoch": 1.8466835246061077,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046562324502949834,
      "loss": 4.9242,
      "mean_token_accuracy": 0.20926232933998107,
      "num_tokens": 41179512.0,
      "step": 23735
    },
    {
      "entropy": 5.600370216369629,
      "epoch": 1.8470725539778252,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004656085196471416,
      "loss": 4.9573,
      "mean_token_accuracy": 0.20923246592283248,
      "num_tokens": 41188109.0,
      "step": 23740
    },
    {
      "entropy": 5.634823417663574,
      "epoch": 1.847461583349543,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046559379137260404,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1925618663430214,
      "num_tokens": 41196708.0,
      "step": 23745
    },
    {
      "entropy": 5.753923177719116,
      "epoch": 1.8478506127212606,
      "grad_norm": 1.203125,
      "learning_rate": 0.000465579060206109,
      "loss": 5.2055,
      "mean_token_accuracy": 0.1920281246304512,
      "num_tokens": 41205013.0,
      "step": 23750
    },
    {
      "entropy": 5.630959320068359,
      "epoch": 1.8482396420929779,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046556432614788015,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2030405282974243,
      "num_tokens": 41213701.0,
      "step": 23755
    },
    {
      "entropy": 5.580765676498413,
      "epoch": 1.8486286714646956,
      "grad_norm": 1.125,
      "learning_rate": 0.0004655495891981409,
      "loss": 4.958,
      "mean_token_accuracy": 0.20130491256713867,
      "num_tokens": 41222450.0,
      "step": 23760
    },
    {
      "entropy": 5.6876341819763185,
      "epoch": 1.8490177008364133,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000465534849357115,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19461113959550858,
      "num_tokens": 41230259.0,
      "step": 23765
    },
    {
      "entropy": 5.657480382919312,
      "epoch": 1.8494067302081307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046552010662502595,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20356453359127044,
      "num_tokens": 41238863.0,
      "step": 23770
    },
    {
      "entropy": 5.622526931762695,
      "epoch": 1.8497957595798482,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004655053610020976,
      "loss": 5.0501,
      "mean_token_accuracy": 0.19909099340438843,
      "num_tokens": 41247260.0,
      "step": 23775
    },
    {
      "entropy": 5.625200653076172,
      "epoch": 1.850184788951566,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004654906124885535,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21325041502714157,
      "num_tokens": 41255475.0,
      "step": 23780
    },
    {
      "entropy": 5.563898134231567,
      "epoch": 1.8505738183232834,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004654758610846176,
      "loss": 4.8514,
      "mean_token_accuracy": 0.21445018649101258,
      "num_tokens": 41263858.0,
      "step": 23785
    },
    {
      "entropy": 5.60622730255127,
      "epoch": 1.8509628476950009,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004654611067905137,
      "loss": 5.0629,
      "mean_token_accuracy": 0.1975773274898529,
      "num_tokens": 41272527.0,
      "step": 23790
    },
    {
      "entropy": 5.655701494216919,
      "epoch": 1.8513518770667186,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046544634960646563,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20171908140182496,
      "num_tokens": 41280647.0,
      "step": 23795
    },
    {
      "entropy": 5.720359182357788,
      "epoch": 1.8517409064384363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004654315895326974,
      "loss": 5.0638,
      "mean_token_accuracy": 0.2066345140337944,
      "num_tokens": 41289462.0,
      "step": 23800
    },
    {
      "entropy": 5.693445158004761,
      "epoch": 1.8521299358101535,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004654168265694328,
      "loss": 4.998,
      "mean_token_accuracy": 0.2058317095041275,
      "num_tokens": 41297178.0,
      "step": 23805
    },
    {
      "entropy": 5.65146770477295,
      "epoch": 1.8525189651818712,
      "grad_norm": 1.140625,
      "learning_rate": 0.000465402060716896,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20386239588260652,
      "num_tokens": 41305405.0,
      "step": 23810
    },
    {
      "entropy": 5.654418516159057,
      "epoch": 1.852907994553589,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000465387291975311,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20021893084049225,
      "num_tokens": 41313330.0,
      "step": 23815
    },
    {
      "entropy": 5.652783679962158,
      "epoch": 1.8532970239253064,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046537252034490185,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20422030985355377,
      "num_tokens": 41321245.0,
      "step": 23820
    },
    {
      "entropy": 5.679351568222046,
      "epoch": 1.8536860532970239,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046535774582589275,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19524348974227906,
      "num_tokens": 41330013.0,
      "step": 23825
    },
    {
      "entropy": 5.711350393295288,
      "epoch": 1.8540750826687415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046534296841850776,
      "loss": 5.0674,
      "mean_token_accuracy": 0.20044457018375397,
      "num_tokens": 41338134.0,
      "step": 23830
    },
    {
      "entropy": 5.6390039920806885,
      "epoch": 1.854464112040459,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046532818812297124,
      "loss": 4.9489,
      "mean_token_accuracy": 0.20367083102464675,
      "num_tokens": 41346693.0,
      "step": 23835
    },
    {
      "entropy": 5.6953917980194095,
      "epoch": 1.8548531414121765,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004653134049395074,
      "loss": 5.1124,
      "mean_token_accuracy": 0.19304307997226716,
      "num_tokens": 41354987.0,
      "step": 23840
    },
    {
      "entropy": 5.625625801086426,
      "epoch": 1.8552421707838942,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004652986188683406,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19381673783063888,
      "num_tokens": 41363703.0,
      "step": 23845
    },
    {
      "entropy": 5.68805422782898,
      "epoch": 1.855631200155612,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046528382990969504,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20043540000915527,
      "num_tokens": 41371983.0,
      "step": 23850
    },
    {
      "entropy": 5.692324018478393,
      "epoch": 1.8560202295273294,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004652690380637953,
      "loss": 5.0846,
      "mean_token_accuracy": 0.19501153528690338,
      "num_tokens": 41380215.0,
      "step": 23855
    },
    {
      "entropy": 5.650061416625976,
      "epoch": 1.8564092588990468,
      "grad_norm": 1.125,
      "learning_rate": 0.0004652542433308657,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20285807847976683,
      "num_tokens": 41388927.0,
      "step": 23860
    },
    {
      "entropy": 5.692520666122436,
      "epoch": 1.8567982882707645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004652394457111309,
      "loss": 5.076,
      "mean_token_accuracy": 0.1914843201637268,
      "num_tokens": 41397823.0,
      "step": 23865
    },
    {
      "entropy": 5.675116395950317,
      "epoch": 1.857187317642482,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046522464520481517,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19928423762321473,
      "num_tokens": 41406322.0,
      "step": 23870
    },
    {
      "entropy": 5.631590127944946,
      "epoch": 1.8575763470141995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046520984181214333,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2019497662782669,
      "num_tokens": 41415118.0,
      "step": 23875
    },
    {
      "entropy": 5.641556310653686,
      "epoch": 1.8579653763859172,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004651950355333398,
      "loss": 5.0837,
      "mean_token_accuracy": 0.19632214605808257,
      "num_tokens": 41424139.0,
      "step": 23880
    },
    {
      "entropy": 5.681218862533569,
      "epoch": 1.8583544057576347,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004651802263686294,
      "loss": 5.1172,
      "mean_token_accuracy": 0.18713202476501464,
      "num_tokens": 41432017.0,
      "step": 23885
    },
    {
      "entropy": 5.6625697135925295,
      "epoch": 1.8587434351293521,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004651654143182367,
      "loss": 5.0317,
      "mean_token_accuracy": 0.19592186957597732,
      "num_tokens": 41440279.0,
      "step": 23890
    },
    {
      "entropy": 5.589608192443848,
      "epoch": 1.8591324645010698,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004651505993823866,
      "loss": 4.933,
      "mean_token_accuracy": 0.2068053662776947,
      "num_tokens": 41449325.0,
      "step": 23895
    },
    {
      "entropy": 5.617387485504151,
      "epoch": 1.8595214938727875,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046513578156130383,
      "loss": 4.9488,
      "mean_token_accuracy": 0.20600970089435577,
      "num_tokens": 41457898.0,
      "step": 23900
    },
    {
      "entropy": 5.598612070083618,
      "epoch": 1.859910523244505,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004651209608552131,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20641370266675949,
      "num_tokens": 41466410.0,
      "step": 23905
    },
    {
      "entropy": 5.720271921157837,
      "epoch": 1.8602995526162225,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046510613726433945,
      "loss": 5.0682,
      "mean_token_accuracy": 0.20251244604587554,
      "num_tokens": 41475061.0,
      "step": 23910
    },
    {
      "entropy": 5.608542966842651,
      "epoch": 1.8606885819879402,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004650913107889078,
      "loss": 5.0546,
      "mean_token_accuracy": 0.19628344029188155,
      "num_tokens": 41483569.0,
      "step": 23915
    },
    {
      "entropy": 5.640122270584106,
      "epoch": 1.8610776113596577,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000465076481429143,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19455556720495223,
      "num_tokens": 41492113.0,
      "step": 23920
    },
    {
      "entropy": 5.677572441101074,
      "epoch": 1.8614666407313751,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004650616491852701,
      "loss": 5.1415,
      "mean_token_accuracy": 0.19140958935022354,
      "num_tokens": 41500722.0,
      "step": 23925
    },
    {
      "entropy": 5.6151947498321535,
      "epoch": 1.8618556701030928,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046504681405751426,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20485322624444963,
      "num_tokens": 41509272.0,
      "step": 23930
    },
    {
      "entropy": 5.691391324996948,
      "epoch": 1.8622446994748103,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046503197604610047,
      "loss": 5.1296,
      "mean_token_accuracy": 0.19357597678899766,
      "num_tokens": 41517379.0,
      "step": 23935
    },
    {
      "entropy": 5.582354211807251,
      "epoch": 1.8626337288465278,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046501713515125393,
      "loss": 5.006,
      "mean_token_accuracy": 0.20674109309911728,
      "num_tokens": 41525737.0,
      "step": 23940
    },
    {
      "entropy": 5.645523548126221,
      "epoch": 1.8630227582182455,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004650022913731998,
      "loss": 5.0405,
      "mean_token_accuracy": 0.19899079948663712,
      "num_tokens": 41535016.0,
      "step": 23945
    },
    {
      "entropy": 5.7155755996704105,
      "epoch": 1.8634117875899632,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046498744471216335,
      "loss": 5.1195,
      "mean_token_accuracy": 0.2013600766658783,
      "num_tokens": 41544253.0,
      "step": 23950
    },
    {
      "entropy": 5.711176109313965,
      "epoch": 1.8638008169616807,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046497259516836974,
      "loss": 5.081,
      "mean_token_accuracy": 0.19150735884904863,
      "num_tokens": 41551907.0,
      "step": 23955
    },
    {
      "entropy": 5.7297204494476315,
      "epoch": 1.8641898463333981,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046495774274204446,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19423890262842178,
      "num_tokens": 41560675.0,
      "step": 23960
    },
    {
      "entropy": 5.687219858169556,
      "epoch": 1.8645788757051158,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004649428874334127,
      "loss": 5.1069,
      "mean_token_accuracy": 0.1934732422232628,
      "num_tokens": 41570256.0,
      "step": 23965
    },
    {
      "entropy": 5.697105121612549,
      "epoch": 1.8649679050768333,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004649280292427,
      "loss": 5.0547,
      "mean_token_accuracy": 0.2003311410546303,
      "num_tokens": 41578990.0,
      "step": 23970
    },
    {
      "entropy": 5.649276208877564,
      "epoch": 1.8653569344485508,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004649131681701318,
      "loss": 5.0669,
      "mean_token_accuracy": 0.1962026461958885,
      "num_tokens": 41587896.0,
      "step": 23975
    },
    {
      "entropy": 5.666079425811768,
      "epoch": 1.8657459638202685,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046489830421593347,
      "loss": 5.0449,
      "mean_token_accuracy": 0.19892151206731795,
      "num_tokens": 41596160.0,
      "step": 23980
    },
    {
      "entropy": 5.686300897598267,
      "epoch": 1.866134993191986,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004648834373803307,
      "loss": 5.042,
      "mean_token_accuracy": 0.19486997723579408,
      "num_tokens": 41605025.0,
      "step": 23985
    },
    {
      "entropy": 5.675208044052124,
      "epoch": 1.8665240225637034,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046486856766354893,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2020961359143257,
      "num_tokens": 41612768.0,
      "step": 23990
    },
    {
      "entropy": 5.640949440002442,
      "epoch": 1.8669130519354211,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046485369506581387,
      "loss": 5.0688,
      "mean_token_accuracy": 0.1968616545200348,
      "num_tokens": 41620892.0,
      "step": 23995
    },
    {
      "entropy": 5.697067308425903,
      "epoch": 1.8673020813071388,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046483881958735124,
      "loss": 5.1228,
      "mean_token_accuracy": 0.19956340044736862,
      "num_tokens": 41629708.0,
      "step": 24000
    },
    {
      "epoch": 1.8673020813071388,
      "eval_entropy": 5.481433009275229,
      "eval_loss": 5.103809833526611,
      "eval_mean_token_accuracy": 0.206297004059373,
      "eval_num_tokens": 41629708.0,
      "eval_runtime": 21.8162,
      "eval_samples_per_second": 1904.918,
      "eval_steps_per_second": 238.126,
      "step": 24000
    },
    {
      "entropy": 5.737192249298095,
      "epoch": 1.8676911106788563,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046482394122838663,
      "loss": 4.9478,
      "mean_token_accuracy": 0.20352384001016616,
      "num_tokens": 41637966.0,
      "step": 24005
    },
    {
      "entropy": 5.622758293151856,
      "epoch": 1.8680801400505738,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046480905998914587,
      "loss": 5.0178,
      "mean_token_accuracy": 0.1953002244234085,
      "num_tokens": 41646726.0,
      "step": 24010
    },
    {
      "entropy": 5.603816175460816,
      "epoch": 1.8684691694222915,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004647941758698547,
      "loss": 5.0083,
      "mean_token_accuracy": 0.19887405037879943,
      "num_tokens": 41655450.0,
      "step": 24015
    },
    {
      "entropy": 5.749894380569458,
      "epoch": 1.868858198794009,
      "grad_norm": 1.15625,
      "learning_rate": 0.000464779288870739,
      "loss": 5.1321,
      "mean_token_accuracy": 0.20005185455083846,
      "num_tokens": 41664295.0,
      "step": 24020
    },
    {
      "entropy": 5.620892333984375,
      "epoch": 1.8692472281657264,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046476439899202464,
      "loss": 4.8535,
      "mean_token_accuracy": 0.20888695269823074,
      "num_tokens": 41672025.0,
      "step": 24025
    },
    {
      "entropy": 5.6129711151123045,
      "epoch": 1.8696362575374441,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046474950623393756,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19992337077856065,
      "num_tokens": 41680360.0,
      "step": 24030
    },
    {
      "entropy": 5.636513376235962,
      "epoch": 1.8700252869091616,
      "grad_norm": 1.125,
      "learning_rate": 0.0004647346105967038,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19602497071027755,
      "num_tokens": 41689865.0,
      "step": 24035
    },
    {
      "entropy": 5.724440431594848,
      "epoch": 1.870414316280879,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004647197120805493,
      "loss": 5.1812,
      "mean_token_accuracy": 0.19200527518987656,
      "num_tokens": 41698583.0,
      "step": 24040
    },
    {
      "entropy": 5.698320150375366,
      "epoch": 1.8708033456525968,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046470481068570013,
      "loss": 5.2314,
      "mean_token_accuracy": 0.18421397656202315,
      "num_tokens": 41708107.0,
      "step": 24045
    },
    {
      "entropy": 5.738505697250366,
      "epoch": 1.8711923750243145,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004646899064123824,
      "loss": 5.1234,
      "mean_token_accuracy": 0.19045273959636688,
      "num_tokens": 41716446.0,
      "step": 24050
    },
    {
      "entropy": 5.673286724090576,
      "epoch": 1.871581404396032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004646749992608223,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20067374855279924,
      "num_tokens": 41726142.0,
      "step": 24055
    },
    {
      "entropy": 5.649875974655151,
      "epoch": 1.8719704337677494,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004646600892312459,
      "loss": 5.0273,
      "mean_token_accuracy": 0.2011725515127182,
      "num_tokens": 41734488.0,
      "step": 24060
    },
    {
      "entropy": 5.662187004089356,
      "epoch": 1.872359463139467,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004646451763238796,
      "loss": 5.0399,
      "mean_token_accuracy": 0.19961638003587723,
      "num_tokens": 41743164.0,
      "step": 24065
    },
    {
      "entropy": 5.667264604568482,
      "epoch": 1.8727484925111846,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004646302605389496,
      "loss": 5.0835,
      "mean_token_accuracy": 0.19702913612127304,
      "num_tokens": 41751994.0,
      "step": 24070
    },
    {
      "entropy": 5.744102144241333,
      "epoch": 1.873137521882902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004646153418766822,
      "loss": 5.1435,
      "mean_token_accuracy": 0.19567607045173646,
      "num_tokens": 41761801.0,
      "step": 24075
    },
    {
      "entropy": 5.7882942199707035,
      "epoch": 1.8735265512546198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046460042033730377,
      "loss": 5.1435,
      "mean_token_accuracy": 0.1981971010565758,
      "num_tokens": 41770771.0,
      "step": 24080
    },
    {
      "entropy": 5.651382493972778,
      "epoch": 1.8739155806263375,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004645854959210408,
      "loss": 5.0771,
      "mean_token_accuracy": 0.19634080529212952,
      "num_tokens": 41779813.0,
      "step": 24085
    },
    {
      "entropy": 5.667339181900024,
      "epoch": 1.8743046099980547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046457056862811956,
      "loss": 5.109,
      "mean_token_accuracy": 0.20441514253616333,
      "num_tokens": 41790076.0,
      "step": 24090
    },
    {
      "entropy": 5.667044305801392,
      "epoch": 1.8746936393697724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004645556384587668,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21150904297828674,
      "num_tokens": 41799448.0,
      "step": 24095
    },
    {
      "entropy": 5.700804090499878,
      "epoch": 1.87508266874149,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004645407054132089,
      "loss": 5.1066,
      "mean_token_accuracy": 0.1968839168548584,
      "num_tokens": 41808117.0,
      "step": 24100
    },
    {
      "entropy": 5.661335563659668,
      "epoch": 1.8754716981132076,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004645257694916725,
      "loss": 4.9962,
      "mean_token_accuracy": 0.20509607642889022,
      "num_tokens": 41816176.0,
      "step": 24105
    },
    {
      "entropy": 5.656861352920532,
      "epoch": 1.875860727484925,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004645108306943842,
      "loss": 5.0823,
      "mean_token_accuracy": 0.19505470097064972,
      "num_tokens": 41825093.0,
      "step": 24110
    },
    {
      "entropy": 5.71199221611023,
      "epoch": 1.8762497568566427,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004644958890215707,
      "loss": 5.1615,
      "mean_token_accuracy": 0.19339849650859833,
      "num_tokens": 41834524.0,
      "step": 24115
    },
    {
      "entropy": 5.628618478775024,
      "epoch": 1.8766387862283602,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004644809444734586,
      "loss": 4.9712,
      "mean_token_accuracy": 0.20600752234458924,
      "num_tokens": 41843188.0,
      "step": 24120
    },
    {
      "entropy": 5.664265489578247,
      "epoch": 1.8770278156000777,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046446599705027487,
      "loss": 5.0742,
      "mean_token_accuracy": 0.19733261317014694,
      "num_tokens": 41852296.0,
      "step": 24125
    },
    {
      "entropy": 5.715221881866455,
      "epoch": 1.8774168449717954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046445104675224607,
      "loss": 5.1118,
      "mean_token_accuracy": 0.19405912160873412,
      "num_tokens": 41861641.0,
      "step": 24130
    },
    {
      "entropy": 5.643478775024414,
      "epoch": 1.877805874343513,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004644360935795993,
      "loss": 5.0287,
      "mean_token_accuracy": 0.19584417939186097,
      "num_tokens": 41870463.0,
      "step": 24135
    },
    {
      "entropy": 5.683660697937012,
      "epoch": 1.8781949037152303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046442113753256126,
      "loss": 5.1071,
      "mean_token_accuracy": 0.19455104768276216,
      "num_tokens": 41878739.0,
      "step": 24140
    },
    {
      "entropy": 5.705801868438721,
      "epoch": 1.878583933086948,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046440617861135905,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19814322590827943,
      "num_tokens": 41887211.0,
      "step": 24145
    },
    {
      "entropy": 5.720928525924682,
      "epoch": 1.8789729624586657,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004643912168162194,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21529816687107087,
      "num_tokens": 41895742.0,
      "step": 24150
    },
    {
      "entropy": 5.659170007705688,
      "epoch": 1.8793619918303832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004643762521473696,
      "loss": 5.0633,
      "mean_token_accuracy": 0.19968766272068023,
      "num_tokens": 41904866.0,
      "step": 24155
    },
    {
      "entropy": 5.693766117095947,
      "epoch": 1.8797510212021007,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004643612846050366,
      "loss": 5.1566,
      "mean_token_accuracy": 0.19215664118528367,
      "num_tokens": 41914237.0,
      "step": 24160
    },
    {
      "entropy": 5.623767948150634,
      "epoch": 1.8801400505738184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046434631418944744,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20290460139513017,
      "num_tokens": 41923070.0,
      "step": 24165
    },
    {
      "entropy": 5.6877528667449955,
      "epoch": 1.8805290799455359,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004643313409008294,
      "loss": 5.0361,
      "mean_token_accuracy": 0.19455904513597488,
      "num_tokens": 41932076.0,
      "step": 24170
    },
    {
      "entropy": 5.670101165771484,
      "epoch": 1.8809181093172533,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046431636473940956,
      "loss": 5.0855,
      "mean_token_accuracy": 0.19343627840280533,
      "num_tokens": 41941169.0,
      "step": 24175
    },
    {
      "entropy": 5.629842233657837,
      "epoch": 1.881307138688971,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004643013857054152,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2065153256058693,
      "num_tokens": 41949499.0,
      "step": 24180
    },
    {
      "entropy": 5.7294573307037355,
      "epoch": 1.8816961680606887,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046428640379907367,
      "loss": 5.1544,
      "mean_token_accuracy": 0.19139719605445862,
      "num_tokens": 41957274.0,
      "step": 24185
    },
    {
      "entropy": 5.63903579711914,
      "epoch": 1.882085197432406,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046427141902061225,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20137618333101273,
      "num_tokens": 41965996.0,
      "step": 24190
    },
    {
      "entropy": 5.685209560394287,
      "epoch": 1.8824742268041237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004642564313702582,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2013673424720764,
      "num_tokens": 41974652.0,
      "step": 24195
    },
    {
      "entropy": 5.687367343902588,
      "epoch": 1.8828632561758414,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046424144084823916,
      "loss": 5.0538,
      "mean_token_accuracy": 0.1953516870737076,
      "num_tokens": 41982699.0,
      "step": 24200
    },
    {
      "entropy": 5.692712688446045,
      "epoch": 1.8832522855475589,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004642264474547824,
      "loss": 5.0037,
      "mean_token_accuracy": 0.2048982560634613,
      "num_tokens": 41991441.0,
      "step": 24205
    },
    {
      "entropy": 5.621842861175537,
      "epoch": 1.8836413149192763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046421145119011556,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20807349681854248,
      "num_tokens": 42000683.0,
      "step": 24210
    },
    {
      "entropy": 5.687556791305542,
      "epoch": 1.884030344290994,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000464196452054466,
      "loss": 5.0792,
      "mean_token_accuracy": 0.2011927545070648,
      "num_tokens": 42009895.0,
      "step": 24215
    },
    {
      "entropy": 5.622824478149414,
      "epoch": 1.8844193736627115,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004641814500480615,
      "loss": 4.9368,
      "mean_token_accuracy": 0.20323821306228637,
      "num_tokens": 42018086.0,
      "step": 24220
    },
    {
      "entropy": 5.6669597148895265,
      "epoch": 1.884808403034429,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004641664451711296,
      "loss": 5.0937,
      "mean_token_accuracy": 0.19554786384105682,
      "num_tokens": 42026959.0,
      "step": 24225
    },
    {
      "entropy": 5.6167539119720455,
      "epoch": 1.8851974324061467,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046415143742389793,
      "loss": 5.0302,
      "mean_token_accuracy": 0.204665932059288,
      "num_tokens": 42035641.0,
      "step": 24230
    },
    {
      "entropy": 5.6266104221344,
      "epoch": 1.8855864617778644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004641364268065943,
      "loss": 5.022,
      "mean_token_accuracy": 0.2042750671505928,
      "num_tokens": 42044565.0,
      "step": 24235
    },
    {
      "entropy": 5.726455736160278,
      "epoch": 1.8859754911495816,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046412141331944644,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1965905100107193,
      "num_tokens": 42053088.0,
      "step": 24240
    },
    {
      "entropy": 5.676874303817749,
      "epoch": 1.8863645205212993,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046410639696268206,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1974782407283783,
      "num_tokens": 42061880.0,
      "step": 24245
    },
    {
      "entropy": 5.600919437408447,
      "epoch": 1.886753549893017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004640913777365292,
      "loss": 5.038,
      "mean_token_accuracy": 0.19703698754310608,
      "num_tokens": 42070767.0,
      "step": 24250
    },
    {
      "entropy": 5.64341459274292,
      "epoch": 1.8871425792647345,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046407635564121565,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20381596982479094,
      "num_tokens": 42080037.0,
      "step": 24255
    },
    {
      "entropy": 5.687326526641845,
      "epoch": 1.887531608636452,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046406133067696936,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20225559920072556,
      "num_tokens": 42088372.0,
      "step": 24260
    },
    {
      "entropy": 5.650011587142944,
      "epoch": 1.8879206380081697,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004640463028440182,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20456052720546722,
      "num_tokens": 42096550.0,
      "step": 24265
    },
    {
      "entropy": 5.65456805229187,
      "epoch": 1.8883096673798871,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004640312721425904,
      "loss": 5.0424,
      "mean_token_accuracy": 0.19975408017635346,
      "num_tokens": 42105660.0,
      "step": 24270
    },
    {
      "entropy": 5.673205852508545,
      "epoch": 1.8886986967516046,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004640162385729139,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20744022428989412,
      "num_tokens": 42114586.0,
      "step": 24275
    },
    {
      "entropy": 5.658952283859253,
      "epoch": 1.8890877261233223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004640012021352168,
      "loss": 5.0498,
      "mean_token_accuracy": 0.20315370410680772,
      "num_tokens": 42123104.0,
      "step": 24280
    },
    {
      "entropy": 5.724965238571167,
      "epoch": 1.88947675549504,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004639861628297273,
      "loss": 5.1624,
      "mean_token_accuracy": 0.19158992022275925,
      "num_tokens": 42132127.0,
      "step": 24285
    },
    {
      "entropy": 5.714989948272705,
      "epoch": 1.8898657848667575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046397112065667354,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19685220420360566,
      "num_tokens": 42141586.0,
      "step": 24290
    },
    {
      "entropy": 5.709984064102173,
      "epoch": 1.890254814238475,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046395607561628387,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19813930988311768,
      "num_tokens": 42151003.0,
      "step": 24295
    },
    {
      "entropy": 5.690791511535645,
      "epoch": 1.8906438436101927,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046394102770878643,
      "loss": 5.0812,
      "mean_token_accuracy": 0.19987884759902955,
      "num_tokens": 42159291.0,
      "step": 24300
    },
    {
      "entropy": 5.729184293746949,
      "epoch": 1.8910328729819101,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046392597693440974,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19339672327041627,
      "num_tokens": 42167925.0,
      "step": 24305
    },
    {
      "entropy": 5.7045917987823485,
      "epoch": 1.8914219023536276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004639109232933819,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20020367205142975,
      "num_tokens": 42175866.0,
      "step": 24310
    },
    {
      "entropy": 5.667725896835327,
      "epoch": 1.8918109317253453,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004638958667859316,
      "loss": 5.0324,
      "mean_token_accuracy": 0.2069874718785286,
      "num_tokens": 42184233.0,
      "step": 24315
    },
    {
      "entropy": 5.637663269042969,
      "epoch": 1.8921999610970628,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004638808074122872,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19847211092710496,
      "num_tokens": 42192972.0,
      "step": 24320
    },
    {
      "entropy": 5.6054713249206545,
      "epoch": 1.8925889904687803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004638657451726771,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20734380483627318,
      "num_tokens": 42201380.0,
      "step": 24325
    },
    {
      "entropy": 5.652434539794922,
      "epoch": 1.892978019840498,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046385068006732995,
      "loss": 5.071,
      "mean_token_accuracy": 0.2010469004511833,
      "num_tokens": 42209276.0,
      "step": 24330
    },
    {
      "entropy": 5.733061456680298,
      "epoch": 1.8933670492122157,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046383561209647437,
      "loss": 5.1979,
      "mean_token_accuracy": 0.19084037393331527,
      "num_tokens": 42218682.0,
      "step": 24335
    },
    {
      "entropy": 5.8094830989837645,
      "epoch": 1.8937560785839331,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046382054126033884,
      "loss": 5.219,
      "mean_token_accuracy": 0.18880195170640945,
      "num_tokens": 42226692.0,
      "step": 24340
    },
    {
      "entropy": 5.750454711914062,
      "epoch": 1.8941451079556506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004638054675591523,
      "loss": 5.1906,
      "mean_token_accuracy": 0.18497735857963563,
      "num_tokens": 42235582.0,
      "step": 24345
    },
    {
      "entropy": 5.759460020065307,
      "epoch": 1.8945341373273683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004637903909931432,
      "loss": 5.0309,
      "mean_token_accuracy": 0.2000938206911087,
      "num_tokens": 42244354.0,
      "step": 24350
    },
    {
      "entropy": 5.685082626342774,
      "epoch": 1.8949231666990858,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004637753115625404,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20758718997240067,
      "num_tokens": 42253062.0,
      "step": 24355
    },
    {
      "entropy": 5.644656610488892,
      "epoch": 1.8953121960708033,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004637602292675726,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19750078320503234,
      "num_tokens": 42262380.0,
      "step": 24360
    },
    {
      "entropy": 5.663568925857544,
      "epoch": 1.895701225442521,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004637451441084689,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2044386699795723,
      "num_tokens": 42270262.0,
      "step": 24365
    },
    {
      "entropy": 5.7212364196777346,
      "epoch": 1.8960902548142384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004637300560854581,
      "loss": 5.091,
      "mean_token_accuracy": 0.20159485191106796,
      "num_tokens": 42278935.0,
      "step": 24370
    },
    {
      "entropy": 5.6866213321685795,
      "epoch": 1.896479284185956,
      "grad_norm": 1.1875,
      "learning_rate": 0.000463714965198769,
      "loss": 5.0101,
      "mean_token_accuracy": 0.19894649386405944,
      "num_tokens": 42287354.0,
      "step": 24375
    },
    {
      "entropy": 5.669860982894898,
      "epoch": 1.8968683135576736,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004636998714486307,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20700916200876235,
      "num_tokens": 42296475.0,
      "step": 24380
    },
    {
      "entropy": 5.709851408004761,
      "epoch": 1.8972573429293913,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046368477483527224,
      "loss": 5.1308,
      "mean_token_accuracy": 0.18853460848331452,
      "num_tokens": 42305306.0,
      "step": 24385
    },
    {
      "entropy": 5.632113027572632,
      "epoch": 1.8976463723011088,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004636696753589226,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19535305202007294,
      "num_tokens": 42314827.0,
      "step": 24390
    },
    {
      "entropy": 5.565219497680664,
      "epoch": 1.8980354016728263,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004636545730198109,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20901987403631211,
      "num_tokens": 42323396.0,
      "step": 24395
    },
    {
      "entropy": 5.709227800369263,
      "epoch": 1.898424431044544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046363946781816643,
      "loss": 5.0493,
      "mean_token_accuracy": 0.19628542065620422,
      "num_tokens": 42331402.0,
      "step": 24400
    },
    {
      "entropy": 5.57415018081665,
      "epoch": 1.8988134604162614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046362435975421816,
      "loss": 4.9885,
      "mean_token_accuracy": 0.2097721889615059,
      "num_tokens": 42339458.0,
      "step": 24405
    },
    {
      "entropy": 5.7077747821807865,
      "epoch": 1.899202489787979,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046360924882819554,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19482335299253464,
      "num_tokens": 42346713.0,
      "step": 24410
    },
    {
      "entropy": 5.661205387115478,
      "epoch": 1.8995915191596966,
      "grad_norm": 1.125,
      "learning_rate": 0.0004635941350403277,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20136691778898239,
      "num_tokens": 42355429.0,
      "step": 24415
    },
    {
      "entropy": 5.588627338409424,
      "epoch": 1.899980548531414,
      "grad_norm": 1.125,
      "learning_rate": 0.0004635790183908442,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20058282762765883,
      "num_tokens": 42364228.0,
      "step": 24420
    },
    {
      "entropy": 5.614884090423584,
      "epoch": 1.9003695779031315,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046356389887997415,
      "loss": 4.9347,
      "mean_token_accuracy": 0.20296090394258498,
      "num_tokens": 42372837.0,
      "step": 24425
    },
    {
      "entropy": 5.643968820571899,
      "epoch": 1.9007586072748492,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046354877650794707,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2135082706809044,
      "num_tokens": 42381792.0,
      "step": 24430
    },
    {
      "entropy": 5.694005632400513,
      "epoch": 1.901147636646567,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046353365127499235,
      "loss": 5.0898,
      "mean_token_accuracy": 0.1965276911854744,
      "num_tokens": 42390650.0,
      "step": 24435
    },
    {
      "entropy": 5.672324275970459,
      "epoch": 1.9015366660182844,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004635185231813396,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20310599505901336,
      "num_tokens": 42399253.0,
      "step": 24440
    },
    {
      "entropy": 5.707872486114502,
      "epoch": 1.901925695390002,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004635033922272183,
      "loss": 5.2488,
      "mean_token_accuracy": 0.18168317675590515,
      "num_tokens": 42407916.0,
      "step": 24445
    },
    {
      "entropy": 5.588143587112427,
      "epoch": 1.9023147247617196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046348825841285813,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2094084233045578,
      "num_tokens": 42416151.0,
      "step": 24450
    },
    {
      "entropy": 5.7098979473114015,
      "epoch": 1.902703754133437,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004634731217384886,
      "loss": 5.1792,
      "mean_token_accuracy": 0.18593926280736922,
      "num_tokens": 42424845.0,
      "step": 24455
    },
    {
      "entropy": 5.794400262832641,
      "epoch": 1.9030927835051545,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004634579822043394,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19536516517400743,
      "num_tokens": 42434022.0,
      "step": 24460
    },
    {
      "entropy": 5.769221544265747,
      "epoch": 1.9034818128768722,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046344283981064035,
      "loss": 5.0969,
      "mean_token_accuracy": 0.19951676577329636,
      "num_tokens": 42443614.0,
      "step": 24465
    },
    {
      "entropy": 5.607101821899414,
      "epoch": 1.90387084224859,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046342769455762114,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2056431621313095,
      "num_tokens": 42452389.0,
      "step": 24470
    },
    {
      "entropy": 5.632593441009521,
      "epoch": 1.9042598716203072,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004634125464455116,
      "loss": 5.0412,
      "mean_token_accuracy": 0.19796082079410554,
      "num_tokens": 42460796.0,
      "step": 24475
    },
    {
      "entropy": 5.698191785812378,
      "epoch": 1.9046489009920249,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046339739547454146,
      "loss": 5.1336,
      "mean_token_accuracy": 0.19448691755533218,
      "num_tokens": 42470108.0,
      "step": 24480
    },
    {
      "entropy": 5.706844472885132,
      "epoch": 1.9050379303637426,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046338224164494074,
      "loss": 5.0789,
      "mean_token_accuracy": 0.19098154157400132,
      "num_tokens": 42478361.0,
      "step": 24485
    },
    {
      "entropy": 5.702779150009155,
      "epoch": 1.90542695973546,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046336708495693933,
      "loss": 5.1329,
      "mean_token_accuracy": 0.19054466336965561,
      "num_tokens": 42486254.0,
      "step": 24490
    },
    {
      "entropy": 5.6431809425354,
      "epoch": 1.9058159891071775,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046335192541076725,
      "loss": 4.9815,
      "mean_token_accuracy": 0.21181270778179168,
      "num_tokens": 42495379.0,
      "step": 24495
    },
    {
      "entropy": 5.676251173019409,
      "epoch": 1.9062050184788952,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004633367630066545,
      "loss": 5.0476,
      "mean_token_accuracy": 0.1975627601146698,
      "num_tokens": 42503777.0,
      "step": 24500
    },
    {
      "entropy": 5.686475515365601,
      "epoch": 1.9065940478506127,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046332159774483116,
      "loss": 5.1221,
      "mean_token_accuracy": 0.19122467637062074,
      "num_tokens": 42512557.0,
      "step": 24505
    },
    {
      "entropy": 5.6799290657043455,
      "epoch": 1.9069830772223302,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004633064296255273,
      "loss": 5.0523,
      "mean_token_accuracy": 0.19714728593826295,
      "num_tokens": 42521245.0,
      "step": 24510
    },
    {
      "entropy": 5.669378042221069,
      "epoch": 1.9073721065940479,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046329125864897307,
      "loss": 5.0009,
      "mean_token_accuracy": 0.2056272640824318,
      "num_tokens": 42529726.0,
      "step": 24515
    },
    {
      "entropy": 5.629570722579956,
      "epoch": 1.9077611359657656,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004632760848153987,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20151450037956237,
      "num_tokens": 42538376.0,
      "step": 24520
    },
    {
      "entropy": 5.659413862228393,
      "epoch": 1.9081501653374828,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004632609081250345,
      "loss": 5.0806,
      "mean_token_accuracy": 0.1968728110194206,
      "num_tokens": 42546977.0,
      "step": 24525
    },
    {
      "entropy": 5.740017032623291,
      "epoch": 1.9085391947092005,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046324572857811054,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1955328568816185,
      "num_tokens": 42555531.0,
      "step": 24530
    },
    {
      "entropy": 5.741284418106079,
      "epoch": 1.9089282240809182,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004632305461748573,
      "loss": 5.0895,
      "mean_token_accuracy": 0.1949363499879837,
      "num_tokens": 42564803.0,
      "step": 24535
    },
    {
      "entropy": 5.702091979980469,
      "epoch": 1.9093172534526357,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046321536091550517,
      "loss": 5.099,
      "mean_token_accuracy": 0.19323325604200364,
      "num_tokens": 42573560.0,
      "step": 24540
    },
    {
      "entropy": 5.687213325500489,
      "epoch": 1.9097062828243532,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004632001728002845,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19506914019584656,
      "num_tokens": 42582177.0,
      "step": 24545
    },
    {
      "entropy": 5.68519926071167,
      "epoch": 1.9100953121960709,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004631849818294257,
      "loss": 5.0589,
      "mean_token_accuracy": 0.19782676100730895,
      "num_tokens": 42590874.0,
      "step": 24550
    },
    {
      "entropy": 5.646425342559814,
      "epoch": 1.9104843415677883,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046316978800315934,
      "loss": 4.9653,
      "mean_token_accuracy": 0.2025894597172737,
      "num_tokens": 42599644.0,
      "step": 24555
    },
    {
      "entropy": 5.674803733825684,
      "epoch": 1.9108733709395058,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000463154591321716,
      "loss": 5.0465,
      "mean_token_accuracy": 0.20220196694135667,
      "num_tokens": 42609207.0,
      "step": 24560
    },
    {
      "entropy": 5.728879785537719,
      "epoch": 1.9112624003112235,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046313939178532624,
      "loss": 5.1434,
      "mean_token_accuracy": 0.18857362419366835,
      "num_tokens": 42618669.0,
      "step": 24565
    },
    {
      "entropy": 5.737988662719727,
      "epoch": 1.9116514296829412,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004631241893942206,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19413752108812332,
      "num_tokens": 42628433.0,
      "step": 24570
    },
    {
      "entropy": 5.666187191009522,
      "epoch": 1.9120404590546585,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046310898414862983,
      "loss": 5.011,
      "mean_token_accuracy": 0.20298488289117814,
      "num_tokens": 42636839.0,
      "step": 24575
    },
    {
      "entropy": 5.613370323181153,
      "epoch": 1.9124294884263762,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004630937760487847,
      "loss": 5.014,
      "mean_token_accuracy": 0.20344187021255494,
      "num_tokens": 42645328.0,
      "step": 24580
    },
    {
      "entropy": 5.686635875701905,
      "epoch": 1.9128185177980939,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004630785650949158,
      "loss": 5.1328,
      "mean_token_accuracy": 0.18789007365703583,
      "num_tokens": 42653883.0,
      "step": 24585
    },
    {
      "entropy": 5.715138053894043,
      "epoch": 1.9132075471698113,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004630633512872541,
      "loss": 5.083,
      "mean_token_accuracy": 0.19617179930210113,
      "num_tokens": 42662364.0,
      "step": 24590
    },
    {
      "entropy": 5.623394632339478,
      "epoch": 1.9135965765415288,
      "grad_norm": 1.125,
      "learning_rate": 0.00046304813462603035,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20719459503889084,
      "num_tokens": 42670424.0,
      "step": 24595
    },
    {
      "entropy": 5.595186424255371,
      "epoch": 1.9139856059132465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004630329151114754,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20762873888015748,
      "num_tokens": 42679494.0,
      "step": 24600
    },
    {
      "entropy": 5.7777525901794435,
      "epoch": 1.914374635284964,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046301769274382034,
      "loss": 5.0714,
      "mean_token_accuracy": 0.2039305329322815,
      "num_tokens": 42688042.0,
      "step": 24605
    },
    {
      "entropy": 5.745438432693481,
      "epoch": 1.9147636646566815,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004630024675232961,
      "loss": 5.1601,
      "mean_token_accuracy": 0.19513863027095796,
      "num_tokens": 42696859.0,
      "step": 24610
    },
    {
      "entropy": 5.711048650741577,
      "epoch": 1.9151526940283992,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046298723945013354,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19480998516082765,
      "num_tokens": 42705291.0,
      "step": 24615
    },
    {
      "entropy": 5.643743085861206,
      "epoch": 1.9155417234001169,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004629720085245639,
      "loss": 5.0757,
      "mean_token_accuracy": 0.19661943167448043,
      "num_tokens": 42713777.0,
      "step": 24620
    },
    {
      "entropy": 5.646687364578247,
      "epoch": 1.9159307527718341,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004629567747468182,
      "loss": 5.131,
      "mean_token_accuracy": 0.19663719981908798,
      "num_tokens": 42722373.0,
      "step": 24625
    },
    {
      "entropy": 5.699374532699585,
      "epoch": 1.9163197821435518,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004629415381171276,
      "loss": 4.9914,
      "mean_token_accuracy": 0.20402838289737701,
      "num_tokens": 42730791.0,
      "step": 24630
    },
    {
      "entropy": 5.695546722412109,
      "epoch": 1.9167088115152695,
      "grad_norm": 1.125,
      "learning_rate": 0.00046292629863572325,
      "loss": 5.0655,
      "mean_token_accuracy": 0.19486727565526962,
      "num_tokens": 42739448.0,
      "step": 24635
    },
    {
      "entropy": 5.689979314804077,
      "epoch": 1.917097840886987,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004629110563028365,
      "loss": 5.0766,
      "mean_token_accuracy": 0.19948528707027435,
      "num_tokens": 42748757.0,
      "step": 24640
    },
    {
      "entropy": 5.673906421661377,
      "epoch": 1.9174868702587045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004628958111186985,
      "loss": 5.011,
      "mean_token_accuracy": 0.19864040613174438,
      "num_tokens": 42757166.0,
      "step": 24645
    },
    {
      "entropy": 5.700387763977051,
      "epoch": 1.9178758996304222,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004628805630835407,
      "loss": 5.0208,
      "mean_token_accuracy": 0.19511144310235978,
      "num_tokens": 42765671.0,
      "step": 24650
    },
    {
      "entropy": 5.716033887863159,
      "epoch": 1.9182649290021396,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004628653121975944,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20266659557819366,
      "num_tokens": 42774499.0,
      "step": 24655
    },
    {
      "entropy": 5.66021089553833,
      "epoch": 1.918653958373857,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000462850058461091,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20045102685689925,
      "num_tokens": 42783237.0,
      "step": 24660
    },
    {
      "entropy": 5.653446817398072,
      "epoch": 1.9190429877455748,
      "grad_norm": 1.25,
      "learning_rate": 0.0004628348018742619,
      "loss": 4.9923,
      "mean_token_accuracy": 0.1994507357478142,
      "num_tokens": 42791217.0,
      "step": 24665
    },
    {
      "entropy": 5.605998802185058,
      "epoch": 1.9194320171172925,
      "grad_norm": 1.125,
      "learning_rate": 0.0004628195424373387,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20538942217826844,
      "num_tokens": 42799885.0,
      "step": 24670
    },
    {
      "entropy": 5.657863283157349,
      "epoch": 1.91982104648901,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046280428015055286,
      "loss": 5.0278,
      "mean_token_accuracy": 0.19827613979578018,
      "num_tokens": 42809141.0,
      "step": 24675
    },
    {
      "entropy": 5.713230037689209,
      "epoch": 1.9202100758607274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046278901501413606,
      "loss": 5.0874,
      "mean_token_accuracy": 0.20186588317155837,
      "num_tokens": 42817809.0,
      "step": 24680
    },
    {
      "entropy": 5.653727626800537,
      "epoch": 1.9205991052324451,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046277374702831983,
      "loss": 4.9915,
      "mean_token_accuracy": 0.19707466661930084,
      "num_tokens": 42826813.0,
      "step": 24685
    },
    {
      "entropy": 5.683885812759399,
      "epoch": 1.9209881346041626,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046275847619333576,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19894684553146363,
      "num_tokens": 42835325.0,
      "step": 24690
    },
    {
      "entropy": 5.735720586776734,
      "epoch": 1.92137716397588,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046274320250941576,
      "loss": 5.1228,
      "mean_token_accuracy": 0.19951232969760896,
      "num_tokens": 42843975.0,
      "step": 24695
    },
    {
      "entropy": 5.672284889221191,
      "epoch": 1.9217661933475978,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046272792597679146,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20752256214618683,
      "num_tokens": 42852146.0,
      "step": 24700
    },
    {
      "entropy": 5.64046688079834,
      "epoch": 1.9221552227193153,
      "grad_norm": 1.125,
      "learning_rate": 0.00046271264659569474,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19573205560445786,
      "num_tokens": 42860958.0,
      "step": 24705
    },
    {
      "entropy": 5.649072551727295,
      "epoch": 1.9225442520910327,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004626973643663573,
      "loss": 4.969,
      "mean_token_accuracy": 0.21015456914901734,
      "num_tokens": 42869281.0,
      "step": 24710
    },
    {
      "entropy": 5.641292905807495,
      "epoch": 1.9229332814627504,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004626820792890112,
      "loss": 5.0681,
      "mean_token_accuracy": 0.20303475856781006,
      "num_tokens": 42878359.0,
      "step": 24715
    },
    {
      "entropy": 5.675673866271973,
      "epoch": 1.9233223108344681,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004626667913638882,
      "loss": 4.8775,
      "mean_token_accuracy": 0.21503861844539643,
      "num_tokens": 42886156.0,
      "step": 24720
    },
    {
      "entropy": 5.585689210891724,
      "epoch": 1.9237113402061856,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004626515005912203,
      "loss": 4.9101,
      "mean_token_accuracy": 0.20475716292858123,
      "num_tokens": 42894970.0,
      "step": 24725
    },
    {
      "entropy": 5.650521898269654,
      "epoch": 1.924100369577903,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004626362069712397,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20292829275131224,
      "num_tokens": 42903319.0,
      "step": 24730
    },
    {
      "entropy": 5.627458143234253,
      "epoch": 1.9244893989496208,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004626209105041782,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19559075683355331,
      "num_tokens": 42911912.0,
      "step": 24735
    },
    {
      "entropy": 5.694595861434936,
      "epoch": 1.9248784283213383,
      "grad_norm": 1.171875,
      "learning_rate": 0.000462605611190268,
      "loss": 5.0108,
      "mean_token_accuracy": 0.2050314337015152,
      "num_tokens": 42920329.0,
      "step": 24740
    },
    {
      "entropy": 5.663406085968018,
      "epoch": 1.9252674576930557,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046259030902974133,
      "loss": 5.0446,
      "mean_token_accuracy": 0.19656941741704942,
      "num_tokens": 42928422.0,
      "step": 24745
    },
    {
      "entropy": 5.657168340682984,
      "epoch": 1.9256564870647734,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004625750040228302,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20125230848789216,
      "num_tokens": 42937016.0,
      "step": 24750
    },
    {
      "entropy": 5.577286338806152,
      "epoch": 1.926045516436491,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046255969616976704,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20298513770103455,
      "num_tokens": 42945209.0,
      "step": 24755
    },
    {
      "entropy": 5.65619478225708,
      "epoch": 1.9264345458082084,
      "grad_norm": 1.140625,
      "learning_rate": 0.000462544385470784,
      "loss": 5.0361,
      "mean_token_accuracy": 0.19371880292892457,
      "num_tokens": 42953798.0,
      "step": 24760
    },
    {
      "entropy": 5.654207515716553,
      "epoch": 1.926823575179926,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004625290719261134,
      "loss": 5.0234,
      "mean_token_accuracy": 0.197957943379879,
      "num_tokens": 42961780.0,
      "step": 24765
    },
    {
      "entropy": 5.659203624725341,
      "epoch": 1.9272126045516438,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004625137555359876,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20338628739118575,
      "num_tokens": 42970087.0,
      "step": 24770
    },
    {
      "entropy": 5.65588231086731,
      "epoch": 1.9276016339233613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046249843630063905,
      "loss": 5.0962,
      "mean_token_accuracy": 0.19749313443899155,
      "num_tokens": 42978732.0,
      "step": 24775
    },
    {
      "entropy": 5.657938718795776,
      "epoch": 1.9279906632950787,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004624831142203001,
      "loss": 5.075,
      "mean_token_accuracy": 0.1990051180124283,
      "num_tokens": 42986899.0,
      "step": 24780
    },
    {
      "entropy": 5.711552286148072,
      "epoch": 1.9283796926667964,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046246778929520346,
      "loss": 5.0331,
      "mean_token_accuracy": 0.19594433754682541,
      "num_tokens": 42995047.0,
      "step": 24785
    },
    {
      "entropy": 5.70040807723999,
      "epoch": 1.928768722038514,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004624524615255814,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20088304281234742,
      "num_tokens": 43004377.0,
      "step": 24790
    },
    {
      "entropy": 5.67579345703125,
      "epoch": 1.9291577514102314,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046243713091166655,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19847961366176606,
      "num_tokens": 43013610.0,
      "step": 24795
    },
    {
      "entropy": 5.667334985733032,
      "epoch": 1.929546780781949,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004624217974536916,
      "loss": 5.0742,
      "mean_token_accuracy": 0.19733272790908812,
      "num_tokens": 43022512.0,
      "step": 24800
    },
    {
      "entropy": 5.58905987739563,
      "epoch": 1.9299358101536666,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046240646115188925,
      "loss": 4.9464,
      "mean_token_accuracy": 0.20724937915802003,
      "num_tokens": 43030588.0,
      "step": 24805
    },
    {
      "entropy": 5.739638090133667,
      "epoch": 1.930324839525384,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004623911220064921,
      "loss": 5.1013,
      "mean_token_accuracy": 0.19450208842754363,
      "num_tokens": 43038717.0,
      "step": 24810
    },
    {
      "entropy": 5.644582843780517,
      "epoch": 1.9307138688971017,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046237578001773297,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2041413053870201,
      "num_tokens": 43046736.0,
      "step": 24815
    },
    {
      "entropy": 5.59522180557251,
      "epoch": 1.9311028982688194,
      "grad_norm": 1.125,
      "learning_rate": 0.00046236043518584454,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2141971230506897,
      "num_tokens": 43054773.0,
      "step": 24820
    },
    {
      "entropy": 5.6454370498657225,
      "epoch": 1.931491927640537,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004623450875110597,
      "loss": 5.048,
      "mean_token_accuracy": 0.19382582157850264,
      "num_tokens": 43062983.0,
      "step": 24825
    },
    {
      "entropy": 5.632285451889038,
      "epoch": 1.9318809570122544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046232973699361147,
      "loss": 4.98,
      "mean_token_accuracy": 0.20601269900798796,
      "num_tokens": 43071161.0,
      "step": 24830
    },
    {
      "entropy": 5.750476884841919,
      "epoch": 1.932269986383972,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004623143836337326,
      "loss": 5.1978,
      "mean_token_accuracy": 0.18581063449382781,
      "num_tokens": 43080514.0,
      "step": 24835
    },
    {
      "entropy": 5.661402797698974,
      "epoch": 1.9326590157556895,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046229902743165607,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2069428488612175,
      "num_tokens": 43089177.0,
      "step": 24840
    },
    {
      "entropy": 5.679691648483276,
      "epoch": 1.933048045127407,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004622836683876149,
      "loss": 5.0973,
      "mean_token_accuracy": 0.19983182698488236,
      "num_tokens": 43097722.0,
      "step": 24845
    },
    {
      "entropy": 5.643684720993042,
      "epoch": 1.9334370744991247,
      "grad_norm": 1.25,
      "learning_rate": 0.0004622683065018422,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20958465784788133,
      "num_tokens": 43106203.0,
      "step": 24850
    },
    {
      "entropy": 5.662094593048096,
      "epoch": 1.9338261038708422,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046225294177457105,
      "loss": 5.0673,
      "mean_token_accuracy": 0.2005910724401474,
      "num_tokens": 43114607.0,
      "step": 24855
    },
    {
      "entropy": 5.764628744125366,
      "epoch": 1.9342151332425597,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046223757420603445,
      "loss": 5.1794,
      "mean_token_accuracy": 0.19267257750034333,
      "num_tokens": 43123179.0,
      "step": 24860
    },
    {
      "entropy": 5.6942400455474855,
      "epoch": 1.9346041626142774,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004622222037964658,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20208489745855332,
      "num_tokens": 43130883.0,
      "step": 24865
    },
    {
      "entropy": 5.541490077972412,
      "epoch": 1.934993191985995,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046220683054609815,
      "loss": 4.7646,
      "mean_token_accuracy": 0.22210109680891038,
      "num_tokens": 43139812.0,
      "step": 24870
    },
    {
      "entropy": 5.661230421066284,
      "epoch": 1.9353822213577125,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046219145445516483,
      "loss": 5.0991,
      "mean_token_accuracy": 0.19494548439979553,
      "num_tokens": 43148756.0,
      "step": 24875
    },
    {
      "entropy": 5.663597154617309,
      "epoch": 1.93577125072943,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046217607552389905,
      "loss": 5.0732,
      "mean_token_accuracy": 0.19117050468921662,
      "num_tokens": 43158131.0,
      "step": 24880
    },
    {
      "entropy": 5.645678949356079,
      "epoch": 1.9361602801011477,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046216069375253435,
      "loss": 4.8888,
      "mean_token_accuracy": 0.20866291970014572,
      "num_tokens": 43166400.0,
      "step": 24885
    },
    {
      "entropy": 5.6859697818756105,
      "epoch": 1.9365493094728652,
      "grad_norm": 1.25,
      "learning_rate": 0.000462145309141304,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2053510293364525,
      "num_tokens": 43175265.0,
      "step": 24890
    },
    {
      "entropy": 5.621823072433472,
      "epoch": 1.9369383388445827,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004621299216904414,
      "loss": 5.0614,
      "mean_token_accuracy": 0.18975958079099656,
      "num_tokens": 43183735.0,
      "step": 24895
    },
    {
      "entropy": 5.731238460540771,
      "epoch": 1.9373273682163004,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046211453140018006,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19664500951766967,
      "num_tokens": 43192217.0,
      "step": 24900
    },
    {
      "entropy": 5.713073444366455,
      "epoch": 1.937716397588018,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004620991382707537,
      "loss": 5.0827,
      "mean_token_accuracy": 0.1919689178466797,
      "num_tokens": 43200241.0,
      "step": 24905
    },
    {
      "entropy": 5.669056940078735,
      "epoch": 1.9381054269597353,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046208374230239556,
      "loss": 5.0423,
      "mean_token_accuracy": 0.1993970215320587,
      "num_tokens": 43209111.0,
      "step": 24910
    },
    {
      "entropy": 5.689865398406982,
      "epoch": 1.938494456331453,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004620683434953394,
      "loss": 5.1112,
      "mean_token_accuracy": 0.19712042212486267,
      "num_tokens": 43219091.0,
      "step": 24915
    },
    {
      "entropy": 5.6750024318695065,
      "epoch": 1.9388834857031707,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046205294184981896,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20292869508266448,
      "num_tokens": 43227140.0,
      "step": 24920
    },
    {
      "entropy": 5.693069553375244,
      "epoch": 1.9392725150748882,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046203753736606787,
      "loss": 5.1266,
      "mean_token_accuracy": 0.19243288785219193,
      "num_tokens": 43236356.0,
      "step": 24925
    },
    {
      "entropy": 5.6377809047698975,
      "epoch": 1.9396615444466057,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004620221300443197,
      "loss": 4.9981,
      "mean_token_accuracy": 0.208082914352417,
      "num_tokens": 43245083.0,
      "step": 24930
    },
    {
      "entropy": 5.70875244140625,
      "epoch": 1.9400505738183234,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004620067198848086,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19584314823150634,
      "num_tokens": 43253042.0,
      "step": 24935
    },
    {
      "entropy": 5.667660427093506,
      "epoch": 1.9404396031900408,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046199130688776805,
      "loss": 5.0756,
      "mean_token_accuracy": 0.19836493581533432,
      "num_tokens": 43261783.0,
      "step": 24940
    },
    {
      "entropy": 5.732879114151001,
      "epoch": 1.9408286325617583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046197589105343204,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20292830616235732,
      "num_tokens": 43270286.0,
      "step": 24945
    },
    {
      "entropy": 5.668138647079468,
      "epoch": 1.941217661933476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004619604723820345,
      "loss": 5.0971,
      "mean_token_accuracy": 0.19891173094511033,
      "num_tokens": 43278402.0,
      "step": 24950
    },
    {
      "entropy": 5.629895973205566,
      "epoch": 1.9416066913051937,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004619450508738094,
      "loss": 5.0655,
      "mean_token_accuracy": 0.20282728374004363,
      "num_tokens": 43286791.0,
      "step": 24955
    },
    {
      "entropy": 5.700758743286133,
      "epoch": 1.941995720676911,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046192962652899056,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19799245297908782,
      "num_tokens": 43295317.0,
      "step": 24960
    },
    {
      "entropy": 5.755083131790161,
      "epoch": 1.9423847500486286,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046191419934781236,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20228767842054368,
      "num_tokens": 43303545.0,
      "step": 24965
    },
    {
      "entropy": 5.712525177001953,
      "epoch": 1.9427737794203463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004618987693305086,
      "loss": 5.1209,
      "mean_token_accuracy": 0.1963297963142395,
      "num_tokens": 43311964.0,
      "step": 24970
    },
    {
      "entropy": 5.692966794967651,
      "epoch": 1.9431628087920638,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004618833364773135,
      "loss": 5.0578,
      "mean_token_accuracy": 0.19987753927707672,
      "num_tokens": 43320137.0,
      "step": 24975
    },
    {
      "entropy": 5.725600528717041,
      "epoch": 1.9435518381637813,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046186790078846127,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19739290475845336,
      "num_tokens": 43328894.0,
      "step": 24980
    },
    {
      "entropy": 5.682323837280274,
      "epoch": 1.943940867535499,
      "grad_norm": 1.125,
      "learning_rate": 0.00046185246226418603,
      "loss": 5.022,
      "mean_token_accuracy": 0.20108211040496826,
      "num_tokens": 43337269.0,
      "step": 24985
    },
    {
      "entropy": 5.670502758026123,
      "epoch": 1.9443298969072165,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046183702090472206,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20041088610887528,
      "num_tokens": 43345838.0,
      "step": 24990
    },
    {
      "entropy": 5.690845775604248,
      "epoch": 1.944718926278934,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004618215767103037,
      "loss": 5.1121,
      "mean_token_accuracy": 0.19926633685827255,
      "num_tokens": 43354395.0,
      "step": 24995
    },
    {
      "entropy": 5.700220537185669,
      "epoch": 1.9451079556506516,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004618061296811653,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20100318789482116,
      "num_tokens": 43363320.0,
      "step": 25000
    },
    {
      "entropy": 5.720243072509765,
      "epoch": 1.9454969850223693,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046179067981754124,
      "loss": 5.0555,
      "mean_token_accuracy": 0.19517334401607514,
      "num_tokens": 43371528.0,
      "step": 25005
    },
    {
      "entropy": 5.701516008377075,
      "epoch": 1.9458860143940866,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046177522711966584,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20645487159490586,
      "num_tokens": 43380177.0,
      "step": 25010
    },
    {
      "entropy": 5.5818675518035885,
      "epoch": 1.9462750437658043,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046175977158777377,
      "loss": 4.909,
      "mean_token_accuracy": 0.21233824640512466,
      "num_tokens": 43387992.0,
      "step": 25015
    },
    {
      "entropy": 5.711573076248169,
      "epoch": 1.946664073137522,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004617443132220993,
      "loss": 5.1331,
      "mean_token_accuracy": 0.1953921690583229,
      "num_tokens": 43397620.0,
      "step": 25020
    },
    {
      "entropy": 5.676930856704712,
      "epoch": 1.9470531025092395,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046172885202287717,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2083299607038498,
      "num_tokens": 43407023.0,
      "step": 25025
    },
    {
      "entropy": 5.674036884307862,
      "epoch": 1.947442131880957,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046171338799034194,
      "loss": 5.0195,
      "mean_token_accuracy": 0.1943240597844124,
      "num_tokens": 43415713.0,
      "step": 25030
    },
    {
      "entropy": 5.617213582992553,
      "epoch": 1.9478311612526746,
      "grad_norm": 1.125,
      "learning_rate": 0.0004616979211247282,
      "loss": 5.0194,
      "mean_token_accuracy": 0.2016974613070488,
      "num_tokens": 43424383.0,
      "step": 25035
    },
    {
      "entropy": 5.658532619476318,
      "epoch": 1.9482201906243921,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046168245142627065,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20131683200597764,
      "num_tokens": 43433090.0,
      "step": 25040
    },
    {
      "entropy": 5.647309255599976,
      "epoch": 1.9486092199961096,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004616669788952041,
      "loss": 5.0008,
      "mean_token_accuracy": 0.19961412250995636,
      "num_tokens": 43442091.0,
      "step": 25045
    },
    {
      "entropy": 5.71430025100708,
      "epoch": 1.9489982493678273,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046165150353176315,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20743487328290938,
      "num_tokens": 43450392.0,
      "step": 25050
    },
    {
      "entropy": 5.643249273300171,
      "epoch": 1.949387278739545,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004616360253361828,
      "loss": 4.991,
      "mean_token_accuracy": 0.20359481424093245,
      "num_tokens": 43459252.0,
      "step": 25055
    },
    {
      "entropy": 5.689338064193725,
      "epoch": 1.9497763081112622,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046162054430869777,
      "loss": 5.1464,
      "mean_token_accuracy": 0.1900984451174736,
      "num_tokens": 43467580.0,
      "step": 25060
    },
    {
      "entropy": 5.806736421585083,
      "epoch": 1.95016533748298,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000461605060449543,
      "loss": 5.1648,
      "mean_token_accuracy": 0.19081234484910964,
      "num_tokens": 43476879.0,
      "step": 25065
    },
    {
      "entropy": 5.751501941680909,
      "epoch": 1.9505543668546976,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046158957375895353,
      "loss": 5.0413,
      "mean_token_accuracy": 0.19494810253381728,
      "num_tokens": 43485881.0,
      "step": 25070
    },
    {
      "entropy": 5.634539079666138,
      "epoch": 1.950943396226415,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004615740842371643,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20470410138368605,
      "num_tokens": 43494598.0,
      "step": 25075
    },
    {
      "entropy": 5.667665481567383,
      "epoch": 1.9513324255981326,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046155859188441023,
      "loss": 5.0651,
      "mean_token_accuracy": 0.1948111116886139,
      "num_tokens": 43504638.0,
      "step": 25080
    },
    {
      "entropy": 5.78444471359253,
      "epoch": 1.9517214549698503,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004615430967009265,
      "loss": 5.1302,
      "mean_token_accuracy": 0.19479121565818786,
      "num_tokens": 43512950.0,
      "step": 25085
    },
    {
      "entropy": 5.721467447280884,
      "epoch": 1.9521104843415678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046152759868694815,
      "loss": 5.1191,
      "mean_token_accuracy": 0.19644711464643477,
      "num_tokens": 43521383.0,
      "step": 25090
    },
    {
      "entropy": 5.577397727966309,
      "epoch": 1.9524995137132852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004615120978427104,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2079143151640892,
      "num_tokens": 43529442.0,
      "step": 25095
    },
    {
      "entropy": 5.690063381195069,
      "epoch": 1.952888543085003,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004614965941684485,
      "loss": 5.0885,
      "mean_token_accuracy": 0.20025897026062012,
      "num_tokens": 43536938.0,
      "step": 25100
    },
    {
      "entropy": 5.64860315322876,
      "epoch": 1.9532775724567206,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004614810876643975,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20373651385307312,
      "num_tokens": 43545337.0,
      "step": 25105
    },
    {
      "entropy": 5.6186055660247805,
      "epoch": 1.953666601828438,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046146557833079286,
      "loss": 4.9441,
      "mean_token_accuracy": 0.20818768739700316,
      "num_tokens": 43553699.0,
      "step": 25110
    },
    {
      "entropy": 5.689172267913818,
      "epoch": 1.9540556312001556,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004614500661678698,
      "loss": 5.1724,
      "mean_token_accuracy": 0.190776264667511,
      "num_tokens": 43562545.0,
      "step": 25115
    },
    {
      "entropy": 5.741392278671265,
      "epoch": 1.9544446605718733,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004614345511758639,
      "loss": 5.0992,
      "mean_token_accuracy": 0.19969851672649383,
      "num_tokens": 43571392.0,
      "step": 25120
    },
    {
      "entropy": 5.668980693817138,
      "epoch": 1.9548336899435907,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046141903335501034,
      "loss": 5.0362,
      "mean_token_accuracy": 0.19697779268026352,
      "num_tokens": 43579864.0,
      "step": 25125
    },
    {
      "entropy": 5.73539457321167,
      "epoch": 1.9552227193153082,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004614035127055447,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1929035782814026,
      "num_tokens": 43588546.0,
      "step": 25130
    },
    {
      "entropy": 5.656674242019653,
      "epoch": 1.955611748687026,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004613879892277024,
      "loss": 4.974,
      "mean_token_accuracy": 0.2053155168890953,
      "num_tokens": 43596582.0,
      "step": 25135
    },
    {
      "entropy": 5.648378467559814,
      "epoch": 1.9560007780587434,
      "grad_norm": 1.1875,
      "learning_rate": 0.000461372462921719,
      "loss": 5.1132,
      "mean_token_accuracy": 0.19592832773923874,
      "num_tokens": 43605727.0,
      "step": 25140
    },
    {
      "entropy": 5.694120025634765,
      "epoch": 1.9563898074304609,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004613569337878302,
      "loss": 5.0784,
      "mean_token_accuracy": 0.19206274449825286,
      "num_tokens": 43614889.0,
      "step": 25145
    },
    {
      "entropy": 5.617065143585205,
      "epoch": 1.9567788368021786,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004613414018262715,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20081593841314316,
      "num_tokens": 43624059.0,
      "step": 25150
    },
    {
      "entropy": 5.756261825561523,
      "epoch": 1.9571678661738963,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004613258670372786,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19828529357910157,
      "num_tokens": 43633824.0,
      "step": 25155
    },
    {
      "entropy": 5.62115797996521,
      "epoch": 1.9575568955456137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004613103294210873,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20632767975330352,
      "num_tokens": 43642010.0,
      "step": 25160
    },
    {
      "entropy": 5.741273021697998,
      "epoch": 1.9579459249173312,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046129478897793323,
      "loss": 5.2046,
      "mean_token_accuracy": 0.19074776619672776,
      "num_tokens": 43650828.0,
      "step": 25165
    },
    {
      "entropy": 5.792493867874145,
      "epoch": 1.958334954289049,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046127924570805236,
      "loss": 5.0808,
      "mean_token_accuracy": 0.19883931279182435,
      "num_tokens": 43659310.0,
      "step": 25170
    },
    {
      "entropy": 5.63452115058899,
      "epoch": 1.9587239836607664,
      "grad_norm": 1.125,
      "learning_rate": 0.0004612636996116803,
      "loss": 5.0042,
      "mean_token_accuracy": 0.19806239008903503,
      "num_tokens": 43668238.0,
      "step": 25175
    },
    {
      "entropy": 5.5849508285522464,
      "epoch": 1.9591130130324839,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004612481506890532,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21630191057920456,
      "num_tokens": 43675748.0,
      "step": 25180
    },
    {
      "entropy": 5.63717212677002,
      "epoch": 1.9595020424042016,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046123259894040677,
      "loss": 5.1038,
      "mean_token_accuracy": 0.19463652968406678,
      "num_tokens": 43685750.0,
      "step": 25185
    },
    {
      "entropy": 5.6520140171051025,
      "epoch": 1.959891071775919,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000461217044365977,
      "loss": 5.0461,
      "mean_token_accuracy": 0.19582537859678267,
      "num_tokens": 43694941.0,
      "step": 25190
    },
    {
      "entropy": 5.766818332672119,
      "epoch": 1.9602801011476365,
      "grad_norm": 1.125,
      "learning_rate": 0.0004612014869660002,
      "loss": 5.1133,
      "mean_token_accuracy": 0.19819205701351167,
      "num_tokens": 43703500.0,
      "step": 25195
    },
    {
      "entropy": 5.76932544708252,
      "epoch": 1.9606691305193542,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046118592674071197,
      "loss": 5.1775,
      "mean_token_accuracy": 0.2000129148364067,
      "num_tokens": 43712235.0,
      "step": 25200
    },
    {
      "entropy": 5.748389530181885,
      "epoch": 1.961058159891072,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004611703636903488,
      "loss": 5.1701,
      "mean_token_accuracy": 0.19067644327878952,
      "num_tokens": 43721047.0,
      "step": 25205
    },
    {
      "entropy": 5.704493999481201,
      "epoch": 1.9614471892627894,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004611547978151466,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20207729786634446,
      "num_tokens": 43729635.0,
      "step": 25210
    },
    {
      "entropy": 5.704960536956787,
      "epoch": 1.9618362186345069,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046113922911534167,
      "loss": 5.04,
      "mean_token_accuracy": 0.19965253472328187,
      "num_tokens": 43737748.0,
      "step": 25215
    },
    {
      "entropy": 5.708689260482788,
      "epoch": 1.9622252480062246,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004611236575911702,
      "loss": 5.015,
      "mean_token_accuracy": 0.21080493032932282,
      "num_tokens": 43745999.0,
      "step": 25220
    },
    {
      "entropy": 5.583837413787842,
      "epoch": 1.962614277377942,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046110808324286844,
      "loss": 4.8823,
      "mean_token_accuracy": 0.2114303082227707,
      "num_tokens": 43754029.0,
      "step": 25225
    },
    {
      "entropy": 5.661626529693604,
      "epoch": 1.9630033067496595,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046109250607067275,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20043956041336058,
      "num_tokens": 43762289.0,
      "step": 25230
    },
    {
      "entropy": 5.665267658233643,
      "epoch": 1.9633923361213772,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004610769260748196,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19312551766633987,
      "num_tokens": 43771479.0,
      "step": 25235
    },
    {
      "entropy": 5.57329363822937,
      "epoch": 1.9637813654930947,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004610613432555451,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20456390529870988,
      "num_tokens": 43779889.0,
      "step": 25240
    },
    {
      "entropy": 5.683544731140136,
      "epoch": 1.9641703948648122,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046104575761308597,
      "loss": 5.0102,
      "mean_token_accuracy": 0.207520093023777,
      "num_tokens": 43788394.0,
      "step": 25245
    },
    {
      "entropy": 5.711502122879028,
      "epoch": 1.9645594242365298,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004610301691476786,
      "loss": 5.0791,
      "mean_token_accuracy": 0.19154712557792664,
      "num_tokens": 43797188.0,
      "step": 25250
    },
    {
      "entropy": 5.754639101028443,
      "epoch": 1.9649484536082475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004610145778595594,
      "loss": 5.1377,
      "mean_token_accuracy": 0.19132084995508195,
      "num_tokens": 43806174.0,
      "step": 25255
    },
    {
      "entropy": 5.635109710693359,
      "epoch": 1.965337482979965,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004609989837489651,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20965223312377929,
      "num_tokens": 43814704.0,
      "step": 25260
    },
    {
      "entropy": 5.609345245361328,
      "epoch": 1.9657265123516825,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046098338681613234,
      "loss": 4.9383,
      "mean_token_accuracy": 0.19789098650217057,
      "num_tokens": 43823122.0,
      "step": 25265
    },
    {
      "entropy": 5.641689586639404,
      "epoch": 1.9661155417234002,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004609677870612976,
      "loss": 5.0179,
      "mean_token_accuracy": 0.205222025513649,
      "num_tokens": 43832281.0,
      "step": 25270
    },
    {
      "entropy": 5.636811208724976,
      "epoch": 1.9665045710951177,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004609521844846978,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20058734714984894,
      "num_tokens": 43840351.0,
      "step": 25275
    },
    {
      "entropy": 5.693105840682984,
      "epoch": 1.9668936004668351,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004609365790865695,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19490966498851775,
      "num_tokens": 43849469.0,
      "step": 25280
    },
    {
      "entropy": 5.719383764266968,
      "epoch": 1.9672826298385528,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046092097086714956,
      "loss": 5.0346,
      "mean_token_accuracy": 0.20038474202156067,
      "num_tokens": 43858013.0,
      "step": 25285
    },
    {
      "entropy": 5.674885416030884,
      "epoch": 1.9676716592102705,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046090535982667486,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20199842154979705,
      "num_tokens": 43867056.0,
      "step": 25290
    },
    {
      "entropy": 5.646392726898194,
      "epoch": 1.9680606885819878,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046088974596538216,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20862494558095931,
      "num_tokens": 43875160.0,
      "step": 25295
    },
    {
      "entropy": 5.613128423690796,
      "epoch": 1.9684497179537055,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004608741292835085,
      "loss": 4.996,
      "mean_token_accuracy": 0.2068445086479187,
      "num_tokens": 43884243.0,
      "step": 25300
    },
    {
      "entropy": 5.64531135559082,
      "epoch": 1.9688387473254232,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004608585097812907,
      "loss": 5.0384,
      "mean_token_accuracy": 0.19544897377490997,
      "num_tokens": 43893378.0,
      "step": 25305
    },
    {
      "entropy": 5.646814250946045,
      "epoch": 1.9692277766971407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004608428874589659,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20425184965133666,
      "num_tokens": 43902095.0,
      "step": 25310
    },
    {
      "entropy": 5.718935012817383,
      "epoch": 1.9696168060688581,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004608272623167711,
      "loss": 5.071,
      "mean_token_accuracy": 0.19806948453187942,
      "num_tokens": 43911452.0,
      "step": 25315
    },
    {
      "entropy": 5.626852941513062,
      "epoch": 1.9700058354405758,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046081163435494335,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2016996592283249,
      "num_tokens": 43920208.0,
      "step": 25320
    },
    {
      "entropy": 5.697553110122681,
      "epoch": 1.9703948648122933,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004607960035737198,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20002957731485366,
      "num_tokens": 43929069.0,
      "step": 25325
    },
    {
      "entropy": 5.773970079421997,
      "epoch": 1.9707838941840108,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004607803699733376,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19867132902145385,
      "num_tokens": 43937444.0,
      "step": 25330
    },
    {
      "entropy": 5.7328794479370115,
      "epoch": 1.9711729235557285,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000460764733554034,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19664835184812546,
      "num_tokens": 43946744.0,
      "step": 25335
    },
    {
      "entropy": 5.7347979068756105,
      "epoch": 1.9715619529274462,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046074909431604623,
      "loss": 5.0911,
      "mean_token_accuracy": 0.19464363306760787,
      "num_tokens": 43954810.0,
      "step": 25340
    },
    {
      "entropy": 5.706856203079224,
      "epoch": 1.9719509822991634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004607334522596116,
      "loss": 5.0722,
      "mean_token_accuracy": 0.1987599402666092,
      "num_tokens": 43963366.0,
      "step": 25345
    },
    {
      "entropy": 5.674899387359619,
      "epoch": 1.9723400116708811,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004607178073849675,
      "loss": 5.012,
      "mean_token_accuracy": 0.202123960852623,
      "num_tokens": 43972170.0,
      "step": 25350
    },
    {
      "entropy": 5.683576679229736,
      "epoch": 1.9727290410425988,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004607021596923512,
      "loss": 5.1094,
      "mean_token_accuracy": 0.193826062977314,
      "num_tokens": 43980743.0,
      "step": 25355
    },
    {
      "entropy": 5.695474910736084,
      "epoch": 1.9731180704143163,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004606865091820003,
      "loss": 5.0002,
      "mean_token_accuracy": 0.1982952445745468,
      "num_tokens": 43990332.0,
      "step": 25360
    },
    {
      "entropy": 5.678173971176148,
      "epoch": 1.9735070997860338,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004606708558541521,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1991661161184311,
      "num_tokens": 43999058.0,
      "step": 25365
    },
    {
      "entropy": 5.681917524337768,
      "epoch": 1.9738961291577515,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004606551997090442,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20519843101501464,
      "num_tokens": 44008310.0,
      "step": 25370
    },
    {
      "entropy": 5.676734590530396,
      "epoch": 1.974285158529469,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004606395407469141,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20411564111709596,
      "num_tokens": 44016818.0,
      "step": 25375
    },
    {
      "entropy": 5.742119836807251,
      "epoch": 1.9746741879011864,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004606238789679995,
      "loss": 5.138,
      "mean_token_accuracy": 0.19193382561206818,
      "num_tokens": 44024755.0,
      "step": 25380
    },
    {
      "entropy": 5.729808568954468,
      "epoch": 1.9750632172729041,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004606082143725381,
      "loss": 5.0825,
      "mean_token_accuracy": 0.2006355881690979,
      "num_tokens": 44033416.0,
      "step": 25385
    },
    {
      "entropy": 5.669182634353637,
      "epoch": 1.9754522466446218,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004605925469607673,
      "loss": 5.0569,
      "mean_token_accuracy": 0.2081127569079399,
      "num_tokens": 44041487.0,
      "step": 25390
    },
    {
      "entropy": 5.688169288635254,
      "epoch": 1.975841276016339,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046057687673292506,
      "loss": 5.0911,
      "mean_token_accuracy": 0.201226170361042,
      "num_tokens": 44050766.0,
      "step": 25395
    },
    {
      "entropy": 5.76194281578064,
      "epoch": 1.9762303053880568,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046056120368924907,
      "loss": 5.1589,
      "mean_token_accuracy": 0.2005791500210762,
      "num_tokens": 44060173.0,
      "step": 25400
    },
    {
      "entropy": 5.682648658752441,
      "epoch": 1.9766193347597745,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004605455278299772,
      "loss": 4.9724,
      "mean_token_accuracy": 0.20670816004276277,
      "num_tokens": 44068472.0,
      "step": 25405
    },
    {
      "entropy": 5.652878284454346,
      "epoch": 1.977008364131492,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004605298491553472,
      "loss": 5.0674,
      "mean_token_accuracy": 0.20026276707649232,
      "num_tokens": 44077209.0,
      "step": 25410
    },
    {
      "entropy": 5.732078313827515,
      "epoch": 1.9773973935032094,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004605141676655971,
      "loss": 5.1742,
      "mean_token_accuracy": 0.19087547212839126,
      "num_tokens": 44086525.0,
      "step": 25415
    },
    {
      "entropy": 5.681312131881714,
      "epoch": 1.9777864228749271,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046049848336096485,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2031294122338295,
      "num_tokens": 44095101.0,
      "step": 25420
    },
    {
      "entropy": 5.669724035263061,
      "epoch": 1.9781754522466446,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004604827962416883,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20618539899587632,
      "num_tokens": 44103469.0,
      "step": 25425
    },
    {
      "entropy": 5.622492456436158,
      "epoch": 1.978564481618362,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004604671063080055,
      "loss": 4.9625,
      "mean_token_accuracy": 0.1973666787147522,
      "num_tokens": 44112161.0,
      "step": 25430
    },
    {
      "entropy": 5.737304449081421,
      "epoch": 1.9789535109900798,
      "grad_norm": 1.125,
      "learning_rate": 0.0004604514135601546,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20211568474769592,
      "num_tokens": 44120452.0,
      "step": 25435
    },
    {
      "entropy": 5.751734638214112,
      "epoch": 1.9793425403617975,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004604357179983736,
      "loss": 5.1177,
      "mean_token_accuracy": 0.19586636275053024,
      "num_tokens": 44129581.0,
      "step": 25440
    },
    {
      "entropy": 5.72521858215332,
      "epoch": 1.9797315697335147,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004604200196229009,
      "loss": 5.0624,
      "mean_token_accuracy": 0.19877135753631592,
      "num_tokens": 44138086.0,
      "step": 25445
    },
    {
      "entropy": 5.6727942943573,
      "epoch": 1.9801205991052324,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004604043184339744,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20334695428609847,
      "num_tokens": 44146549.0,
      "step": 25450
    },
    {
      "entropy": 5.775639772415161,
      "epoch": 1.9805096284769501,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004603886144318325,
      "loss": 5.0825,
      "mean_token_accuracy": 0.19425911605358123,
      "num_tokens": 44156142.0,
      "step": 25455
    },
    {
      "entropy": 5.705142545700073,
      "epoch": 1.9808986578486676,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046037290761671346,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20362311899662017,
      "num_tokens": 44165023.0,
      "step": 25460
    },
    {
      "entropy": 5.69218373298645,
      "epoch": 1.981287687220385,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004603571979888556,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19777097404003144,
      "num_tokens": 44174034.0,
      "step": 25465
    },
    {
      "entropy": 5.725568151473999,
      "epoch": 1.9816767165921028,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004603414855484973,
      "loss": 5.1264,
      "mean_token_accuracy": 0.19230873435735701,
      "num_tokens": 44183072.0,
      "step": 25470
    },
    {
      "entropy": 5.703180027008057,
      "epoch": 1.9820657459638202,
      "grad_norm": 1.0625,
      "learning_rate": 0.000460325770295877,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19481797963380815,
      "num_tokens": 44192181.0,
      "step": 25475
    },
    {
      "entropy": 5.682953500747681,
      "epoch": 1.9824547753355377,
      "grad_norm": 1.125,
      "learning_rate": 0.00046031005223123297,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2032751262187958,
      "num_tokens": 44200967.0,
      "step": 25480
    },
    {
      "entropy": 5.7034485816955565,
      "epoch": 1.9828438047072554,
      "grad_norm": 1.109375,
      "learning_rate": 0.000460294331354804,
      "loss": 5.0457,
      "mean_token_accuracy": 0.2021653413772583,
      "num_tokens": 44209432.0,
      "step": 25485
    },
    {
      "entropy": 5.677782201766968,
      "epoch": 1.983232834078973,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004602786076668283,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2057213380932808,
      "num_tokens": 44218503.0,
      "step": 25490
    },
    {
      "entropy": 5.653190803527832,
      "epoch": 1.9836218634506906,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046026288116754477,
      "loss": 4.9796,
      "mean_token_accuracy": 0.20599444210529327,
      "num_tokens": 44227164.0,
      "step": 25495
    },
    {
      "entropy": 5.6542541027069095,
      "epoch": 1.984010892822408,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004602471518571919,
      "loss": 5.0419,
      "mean_token_accuracy": 0.20098648816347123,
      "num_tokens": 44235950.0,
      "step": 25500
    },
    {
      "entropy": 5.667557954788208,
      "epoch": 1.9843999221941258,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004602314197360083,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2030610114336014,
      "num_tokens": 44244098.0,
      "step": 25505
    },
    {
      "entropy": 5.6968665599823,
      "epoch": 1.9847889515658432,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004602156848042328,
      "loss": 5.0953,
      "mean_token_accuracy": 0.19668178111314774,
      "num_tokens": 44252798.0,
      "step": 25510
    },
    {
      "entropy": 5.712417984008789,
      "epoch": 1.9851779809375607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000460199947062104,
      "loss": 5.0938,
      "mean_token_accuracy": 0.19697902202606202,
      "num_tokens": 44261662.0,
      "step": 25515
    },
    {
      "entropy": 5.744088506698608,
      "epoch": 1.9855670103092784,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046018420650986074,
      "loss": 5.1321,
      "mean_token_accuracy": 0.19806327819824218,
      "num_tokens": 44270480.0,
      "step": 25520
    },
    {
      "entropy": 5.739399003982544,
      "epoch": 1.9859560396809959,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000460168463147742,
      "loss": 5.0468,
      "mean_token_accuracy": 0.20030994415283204,
      "num_tokens": 44278673.0,
      "step": 25525
    },
    {
      "entropy": 5.577852773666382,
      "epoch": 1.9863450690527134,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004601527169759865,
      "loss": 4.9341,
      "mean_token_accuracy": 0.20676348656415938,
      "num_tokens": 44287465.0,
      "step": 25530
    },
    {
      "entropy": 5.666701173782348,
      "epoch": 1.986734098424431,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004601369679948333,
      "loss": 5.1279,
      "mean_token_accuracy": 0.19267341792583464,
      "num_tokens": 44296204.0,
      "step": 25535
    },
    {
      "entropy": 5.803724479675293,
      "epoch": 1.9871231277961487,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046012121620452127,
      "loss": 5.0883,
      "mean_token_accuracy": 0.19376621395349503,
      "num_tokens": 44305338.0,
      "step": 25540
    },
    {
      "entropy": 5.6931699275970455,
      "epoch": 1.9875121571678662,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004601054616052893,
      "loss": 4.9812,
      "mean_token_accuracy": 0.2055448517203331,
      "num_tokens": 44312960.0,
      "step": 25545
    },
    {
      "entropy": 5.660006284713745,
      "epoch": 1.9879011865395837,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046008970419737674,
      "loss": 5.007,
      "mean_token_accuracy": 0.20097713470458983,
      "num_tokens": 44321367.0,
      "step": 25550
    },
    {
      "entropy": 5.639643716812134,
      "epoch": 1.9882902159113014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004600739439810225,
      "loss": 5.0733,
      "mean_token_accuracy": 0.19710378348827362,
      "num_tokens": 44330294.0,
      "step": 25555
    },
    {
      "entropy": 5.739472913742065,
      "epoch": 1.9886792452830189,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046005818095646564,
      "loss": 5.1011,
      "mean_token_accuracy": 0.1988327071070671,
      "num_tokens": 44338639.0,
      "step": 25560
    },
    {
      "entropy": 5.639293813705445,
      "epoch": 1.9890682746547363,
      "grad_norm": 1.125,
      "learning_rate": 0.00046004241512394544,
      "loss": 4.9685,
      "mean_token_accuracy": 0.20167377442121506,
      "num_tokens": 44346911.0,
      "step": 25565
    },
    {
      "entropy": 5.655694532394409,
      "epoch": 1.989457304026454,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004600266464837012,
      "loss": 5.0556,
      "mean_token_accuracy": 0.20089794397354127,
      "num_tokens": 44355580.0,
      "step": 25570
    },
    {
      "entropy": 5.708679628372193,
      "epoch": 1.9898463333981715,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004600108750359721,
      "loss": 5.0917,
      "mean_token_accuracy": 0.19418604969978331,
      "num_tokens": 44363465.0,
      "step": 25575
    },
    {
      "entropy": 5.678712511062622,
      "epoch": 1.990235362769889,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045999510078099736,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21009757816791536,
      "num_tokens": 44372797.0,
      "step": 25580
    },
    {
      "entropy": 5.647272634506225,
      "epoch": 1.9906243921416067,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045997932371901645,
      "loss": 5.0359,
      "mean_token_accuracy": 0.1987574502825737,
      "num_tokens": 44381537.0,
      "step": 25585
    },
    {
      "entropy": 5.674674224853516,
      "epoch": 1.9910134215133244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004599635438502687,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20160763710737228,
      "num_tokens": 44390170.0,
      "step": 25590
    },
    {
      "entropy": 5.728264617919922,
      "epoch": 1.9914024508850419,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045994776117499363,
      "loss": 5.1166,
      "mean_token_accuracy": 0.1958017334342003,
      "num_tokens": 44398545.0,
      "step": 25595
    },
    {
      "entropy": 5.687517499923706,
      "epoch": 1.9917914802567593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045993197569343063,
      "loss": 5.0776,
      "mean_token_accuracy": 0.1949489086866379,
      "num_tokens": 44407587.0,
      "step": 25600
    },
    {
      "entropy": 5.678821039199829,
      "epoch": 1.992180509628477,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045991618740581926,
      "loss": 4.9741,
      "mean_token_accuracy": 0.19748820960521699,
      "num_tokens": 44416005.0,
      "step": 25605
    },
    {
      "entropy": 5.655780458450318,
      "epoch": 1.9925695390001945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004599003963123991,
      "loss": 4.978,
      "mean_token_accuracy": 0.20734387934207915,
      "num_tokens": 44425450.0,
      "step": 25610
    },
    {
      "entropy": 5.6656698226928714,
      "epoch": 1.992958568371912,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045988460241340966,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2067521408200264,
      "num_tokens": 44434657.0,
      "step": 25615
    },
    {
      "entropy": 5.747182321548462,
      "epoch": 1.9933475977436297,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045986880570909075,
      "loss": 5.0579,
      "mean_token_accuracy": 0.19135167300701142,
      "num_tokens": 44443160.0,
      "step": 25620
    },
    {
      "entropy": 5.684082746505737,
      "epoch": 1.9937366271153472,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045985300619968186,
      "loss": 5.0125,
      "mean_token_accuracy": 0.19804889410734178,
      "num_tokens": 44452097.0,
      "step": 25625
    },
    {
      "entropy": 5.6795447826385494,
      "epoch": 1.9941256564870646,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045983720388542286,
      "loss": 4.9927,
      "mean_token_accuracy": 0.1971436470746994,
      "num_tokens": 44459674.0,
      "step": 25630
    },
    {
      "entropy": 5.613994216918945,
      "epoch": 1.9945146858587823,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004598213987665535,
      "loss": 5.0778,
      "mean_token_accuracy": 0.2023503676056862,
      "num_tokens": 44468315.0,
      "step": 25635
    },
    {
      "entropy": 5.756473207473755,
      "epoch": 1.9949037152305,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045980559084331354,
      "loss": 5.092,
      "mean_token_accuracy": 0.20182113498449325,
      "num_tokens": 44477436.0,
      "step": 25640
    },
    {
      "entropy": 5.787140369415283,
      "epoch": 1.9952927446022175,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000459789780115943,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19827466756105422,
      "num_tokens": 44485518.0,
      "step": 25645
    },
    {
      "entropy": 5.659876966476441,
      "epoch": 1.995681773973935,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045977396658468156,
      "loss": 4.9911,
      "mean_token_accuracy": 0.19837686866521836,
      "num_tokens": 44493734.0,
      "step": 25650
    },
    {
      "entropy": 5.640723371505738,
      "epoch": 1.9960708033456527,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004597581502497693,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20193118900060653,
      "num_tokens": 44502615.0,
      "step": 25655
    },
    {
      "entropy": 5.76452145576477,
      "epoch": 1.9964598327173702,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004597423311114462,
      "loss": 5.1275,
      "mean_token_accuracy": 0.19727417230606079,
      "num_tokens": 44511792.0,
      "step": 25660
    },
    {
      "entropy": 5.7669273853302006,
      "epoch": 1.9968488620890876,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004597265091699522,
      "loss": 5.0752,
      "mean_token_accuracy": 0.19535187780857086,
      "num_tokens": 44520627.0,
      "step": 25665
    },
    {
      "entropy": 5.6676344871521,
      "epoch": 1.9972378914608053,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004597106844255276,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20242749005556107,
      "num_tokens": 44528844.0,
      "step": 25670
    },
    {
      "entropy": 5.638690090179443,
      "epoch": 1.9976269208325228,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045969485687841227,
      "loss": 5.1196,
      "mean_token_accuracy": 0.19523014426231383,
      "num_tokens": 44537602.0,
      "step": 25675
    },
    {
      "entropy": 5.710576677322388,
      "epoch": 1.9980159502042403,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045967902652884645,
      "loss": 5.0803,
      "mean_token_accuracy": 0.19598342925310136,
      "num_tokens": 44546261.0,
      "step": 25680
    },
    {
      "entropy": 5.64246826171875,
      "epoch": 1.998404979575958,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004596631933770704,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19503631293773652,
      "num_tokens": 44554291.0,
      "step": 25685
    },
    {
      "entropy": 5.652225160598755,
      "epoch": 1.9987940089476757,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045964735742332427,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20676405727863312,
      "num_tokens": 44562355.0,
      "step": 25690
    },
    {
      "entropy": 5.6939393997192385,
      "epoch": 1.9991830383193931,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004596315186678484,
      "loss": 5.0025,
      "mean_token_accuracy": 0.2014622151851654,
      "num_tokens": 44570517.0,
      "step": 25695
    },
    {
      "entropy": 5.650333070755005,
      "epoch": 1.9995720676911106,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045961567711088307,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20008459389209748,
      "num_tokens": 44578871.0,
      "step": 25700
    },
    {
      "entropy": 5.672059011459351,
      "epoch": 1.9999610970628283,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045959983275266873,
      "loss": 5.0541,
      "mean_token_accuracy": 0.20089416354894638,
      "num_tokens": 44588081.0,
      "step": 25705
    },
    {
      "entropy": 5.783054192860921,
      "epoch": 2.000311223497374,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045958398559344573,
      "loss": 5.1233,
      "mean_token_accuracy": 0.1965047601196501,
      "num_tokens": 44596153.0,
      "step": 25710
    },
    {
      "entropy": 5.683004331588745,
      "epoch": 2.0007002528690916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045956813563345454,
      "loss": 4.9366,
      "mean_token_accuracy": 0.20220261961221694,
      "num_tokens": 44604760.0,
      "step": 25715
    },
    {
      "entropy": 5.607852363586426,
      "epoch": 2.0010892822408093,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004595522828729357,
      "loss": 4.8759,
      "mean_token_accuracy": 0.21571802198886872,
      "num_tokens": 44613251.0,
      "step": 25720
    },
    {
      "entropy": 5.617628574371338,
      "epoch": 2.0014783116125265,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004595364273121296,
      "loss": 4.9771,
      "mean_token_accuracy": 0.214677132666111,
      "num_tokens": 44621914.0,
      "step": 25725
    },
    {
      "entropy": 5.501031684875488,
      "epoch": 2.0018673409842442,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004595205689512771,
      "loss": 4.8724,
      "mean_token_accuracy": 0.21446221470832824,
      "num_tokens": 44630096.0,
      "step": 25730
    },
    {
      "entropy": 5.6222762107849125,
      "epoch": 2.002256370355962,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045950470779061855,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20586709827184677,
      "num_tokens": 44638175.0,
      "step": 25735
    },
    {
      "entropy": 5.720702457427978,
      "epoch": 2.0026453997276796,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045948884383039475,
      "loss": 4.9865,
      "mean_token_accuracy": 0.1971746191382408,
      "num_tokens": 44646970.0,
      "step": 25740
    },
    {
      "entropy": 5.72327938079834,
      "epoch": 2.003034429099397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045947297707084637,
      "loss": 5.0744,
      "mean_token_accuracy": 0.1970928743481636,
      "num_tokens": 44655789.0,
      "step": 25745
    },
    {
      "entropy": 5.6321484565734865,
      "epoch": 2.0034234584711146,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004594571075122142,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2137918934226036,
      "num_tokens": 44664190.0,
      "step": 25750
    },
    {
      "entropy": 5.61806001663208,
      "epoch": 2.0038124878428323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045944123515473905,
      "loss": 4.9553,
      "mean_token_accuracy": 0.205129536986351,
      "num_tokens": 44673295.0,
      "step": 25755
    },
    {
      "entropy": 5.729433917999268,
      "epoch": 2.0042015172145495,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045942535999866175,
      "loss": 5.0542,
      "mean_token_accuracy": 0.20010913908481598,
      "num_tokens": 44681890.0,
      "step": 25760
    },
    {
      "entropy": 5.734658145904541,
      "epoch": 2.0045905465862672,
      "grad_norm": 1.125,
      "learning_rate": 0.0004594094820442231,
      "loss": 5.0436,
      "mean_token_accuracy": 0.19812644720077516,
      "num_tokens": 44690915.0,
      "step": 25765
    },
    {
      "entropy": 5.711312103271484,
      "epoch": 2.004979575957985,
      "grad_norm": 1.1875,
      "learning_rate": 0.000459393601291664,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20076711475849152,
      "num_tokens": 44700075.0,
      "step": 25770
    },
    {
      "entropy": 5.603235054016113,
      "epoch": 2.005368605329702,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045937771774122563,
      "loss": 4.8008,
      "mean_token_accuracy": 0.2216716319322586,
      "num_tokens": 44708732.0,
      "step": 25775
    },
    {
      "entropy": 5.653909206390381,
      "epoch": 2.00575763470142,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004593618313931488,
      "loss": 4.9398,
      "mean_token_accuracy": 0.20578214824199675,
      "num_tokens": 44717019.0,
      "step": 25780
    },
    {
      "entropy": 5.688849020004272,
      "epoch": 2.0061466640731376,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045934594224767463,
      "loss": 4.984,
      "mean_token_accuracy": 0.2104041337966919,
      "num_tokens": 44725359.0,
      "step": 25785
    },
    {
      "entropy": 5.700521039962768,
      "epoch": 2.0065356934448553,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045933005030504424,
      "loss": 4.9708,
      "mean_token_accuracy": 0.20386857837438582,
      "num_tokens": 44733913.0,
      "step": 25790
    },
    {
      "entropy": 5.708501863479614,
      "epoch": 2.0069247228165725,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045931415556549863,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1920447587966919,
      "num_tokens": 44743019.0,
      "step": 25795
    },
    {
      "entropy": 5.677588701248169,
      "epoch": 2.0073137521882902,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004592982580292792,
      "loss": 5.0236,
      "mean_token_accuracy": 0.2042313411831856,
      "num_tokens": 44753140.0,
      "step": 25800
    },
    {
      "entropy": 5.687279891967774,
      "epoch": 2.007702781560008,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045928235769662697,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2045952707529068,
      "num_tokens": 44761470.0,
      "step": 25805
    },
    {
      "entropy": 5.565837764739991,
      "epoch": 2.008091810931725,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045926645456778327,
      "loss": 4.8934,
      "mean_token_accuracy": 0.21055896282196046,
      "num_tokens": 44770570.0,
      "step": 25810
    },
    {
      "entropy": 5.687198877334595,
      "epoch": 2.008480840303443,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045925054864298946,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20308766514062881,
      "num_tokens": 44779551.0,
      "step": 25815
    },
    {
      "entropy": 5.660571765899658,
      "epoch": 2.0088698696751606,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045923463992248684,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20935762226581572,
      "num_tokens": 44787633.0,
      "step": 25820
    },
    {
      "entropy": 5.665345859527588,
      "epoch": 2.009258899046878,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045921872840651675,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2036691889166832,
      "num_tokens": 44796030.0,
      "step": 25825
    },
    {
      "entropy": 5.705021667480469,
      "epoch": 2.0096479284185955,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045920281409532064,
      "loss": 5.048,
      "mean_token_accuracy": 0.2081379696726799,
      "num_tokens": 44805055.0,
      "step": 25830
    },
    {
      "entropy": 5.594717407226563,
      "epoch": 2.010036957790313,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004591868969891401,
      "loss": 4.9235,
      "mean_token_accuracy": 0.20753616392612456,
      "num_tokens": 44813767.0,
      "step": 25835
    },
    {
      "entropy": 5.7365845203399655,
      "epoch": 2.010425987162031,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004591709770882165,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2017764151096344,
      "num_tokens": 44821955.0,
      "step": 25840
    },
    {
      "entropy": 5.71373143196106,
      "epoch": 2.010815016533748,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004591550543927915,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20059753954410553,
      "num_tokens": 44830565.0,
      "step": 25845
    },
    {
      "entropy": 5.662853240966797,
      "epoch": 2.011204045905466,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004591391289031067,
      "loss": 4.9339,
      "mean_token_accuracy": 0.207343128323555,
      "num_tokens": 44839224.0,
      "step": 25850
    },
    {
      "entropy": 5.632823705673218,
      "epoch": 2.0115930752771836,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004591232006194036,
      "loss": 4.8668,
      "mean_token_accuracy": 0.21025255024433137,
      "num_tokens": 44847756.0,
      "step": 25855
    },
    {
      "entropy": 5.712333869934082,
      "epoch": 2.011982104648901,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045910726954192404,
      "loss": 5.0559,
      "mean_token_accuracy": 0.2033000111579895,
      "num_tokens": 44855739.0,
      "step": 25860
    },
    {
      "entropy": 5.68774676322937,
      "epoch": 2.0123711340206185,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004590913356709097,
      "loss": 5.0319,
      "mean_token_accuracy": 0.19616121500730516,
      "num_tokens": 44864402.0,
      "step": 25865
    },
    {
      "entropy": 5.682756328582764,
      "epoch": 2.012760163392336,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045907539900660237,
      "loss": 5.0176,
      "mean_token_accuracy": 0.20379090905189515,
      "num_tokens": 44873369.0,
      "step": 25870
    },
    {
      "entropy": 5.668654918670654,
      "epoch": 2.013149192764054,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004590594595492438,
      "loss": 5.0366,
      "mean_token_accuracy": 0.19428878724575044,
      "num_tokens": 44882464.0,
      "step": 25875
    },
    {
      "entropy": 5.649906730651855,
      "epoch": 2.013538222135771,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045904351729907593,
      "loss": 5.0353,
      "mean_token_accuracy": 0.19852353483438492,
      "num_tokens": 44891729.0,
      "step": 25880
    },
    {
      "entropy": 5.747977924346924,
      "epoch": 2.013927251507489,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045902757225634066,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20086934715509414,
      "num_tokens": 44901752.0,
      "step": 25885
    },
    {
      "entropy": 5.691731739044189,
      "epoch": 2.0143162808792066,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004590116244212799,
      "loss": 4.9836,
      "mean_token_accuracy": 0.20670763105154039,
      "num_tokens": 44909899.0,
      "step": 25890
    },
    {
      "entropy": 5.720140504837036,
      "epoch": 2.014705310250924,
      "grad_norm": 1.25,
      "learning_rate": 0.0004589956737941355,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20399002134799957,
      "num_tokens": 44918472.0,
      "step": 25895
    },
    {
      "entropy": 5.669563579559326,
      "epoch": 2.0150943396226415,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004589797203751497,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21971460282802582,
      "num_tokens": 44926799.0,
      "step": 25900
    },
    {
      "entropy": 5.56992392539978,
      "epoch": 2.015483368994359,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004589637641645645,
      "loss": 4.922,
      "mean_token_accuracy": 0.20980791747570038,
      "num_tokens": 44935867.0,
      "step": 25905
    },
    {
      "entropy": 5.659797096252442,
      "epoch": 2.0158723983660765,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045894780516262193,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20305085927248,
      "num_tokens": 44944705.0,
      "step": 25910
    },
    {
      "entropy": 5.601522207260132,
      "epoch": 2.016261427737794,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004589318433695642,
      "loss": 4.8562,
      "mean_token_accuracy": 0.2161797732114792,
      "num_tokens": 44954301.0,
      "step": 25915
    },
    {
      "entropy": 5.6992229461669925,
      "epoch": 2.016650457109512,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004589158787856336,
      "loss": 4.9237,
      "mean_token_accuracy": 0.20532822608947754,
      "num_tokens": 44963005.0,
      "step": 25920
    },
    {
      "entropy": 5.645299911499023,
      "epoch": 2.0170394864812295,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004588999114110721,
      "loss": 4.959,
      "mean_token_accuracy": 0.20448867231607437,
      "num_tokens": 44970935.0,
      "step": 25925
    },
    {
      "entropy": 5.666358900070191,
      "epoch": 2.017428515852947,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004588839412461223,
      "loss": 4.98,
      "mean_token_accuracy": 0.20308127403259277,
      "num_tokens": 44979915.0,
      "step": 25930
    },
    {
      "entropy": 5.716323614120483,
      "epoch": 2.0178175452246645,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004588679682910264,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20117948055267335,
      "num_tokens": 44988934.0,
      "step": 25935
    },
    {
      "entropy": 5.775421524047852,
      "epoch": 2.018206574596382,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004588519925460266,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19767819344997406,
      "num_tokens": 44997578.0,
      "step": 25940
    },
    {
      "entropy": 5.715643978118896,
      "epoch": 2.0185956039680995,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004588360140113655,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20278383642435074,
      "num_tokens": 45005806.0,
      "step": 25945
    },
    {
      "entropy": 5.657835483551025,
      "epoch": 2.018984633339817,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004588200326872855,
      "loss": 4.9732,
      "mean_token_accuracy": 0.20293842554092406,
      "num_tokens": 45013948.0,
      "step": 25950
    },
    {
      "entropy": 5.66178331375122,
      "epoch": 2.019373662711535,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004588040485740291,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20926092565059662,
      "num_tokens": 45022636.0,
      "step": 25955
    },
    {
      "entropy": 5.77303032875061,
      "epoch": 2.019762692083252,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004587880616718387,
      "loss": 5.0768,
      "mean_token_accuracy": 0.19404577910900117,
      "num_tokens": 45031194.0,
      "step": 25960
    },
    {
      "entropy": 5.75410008430481,
      "epoch": 2.02015172145497,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004587720719809572,
      "loss": 5.0427,
      "mean_token_accuracy": 0.1984744369983673,
      "num_tokens": 45039436.0,
      "step": 25965
    },
    {
      "entropy": 5.634564828872681,
      "epoch": 2.0205407508266875,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004587560795016269,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21031382232904433,
      "num_tokens": 45048173.0,
      "step": 25970
    },
    {
      "entropy": 5.68442702293396,
      "epoch": 2.020929780198405,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004587400842340905,
      "loss": 5.0788,
      "mean_token_accuracy": 0.19508014917373656,
      "num_tokens": 45057896.0,
      "step": 25975
    },
    {
      "entropy": 5.723361253738403,
      "epoch": 2.0213188095701224,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045872408617859083,
      "loss": 5.0148,
      "mean_token_accuracy": 0.1973769560456276,
      "num_tokens": 45066451.0,
      "step": 25980
    },
    {
      "entropy": 5.621495103836059,
      "epoch": 2.02170783894184,
      "grad_norm": 1.125,
      "learning_rate": 0.00045870808533537066,
      "loss": 4.8929,
      "mean_token_accuracy": 0.20970305502414704,
      "num_tokens": 45075143.0,
      "step": 25985
    },
    {
      "entropy": 5.706840467453003,
      "epoch": 2.022096868313558,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045869208170467256,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20515755265951158,
      "num_tokens": 45083370.0,
      "step": 25990
    },
    {
      "entropy": 5.668290662765503,
      "epoch": 2.022485897685275,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004586760752867396,
      "loss": 4.9739,
      "mean_token_accuracy": 0.20159512907266616,
      "num_tokens": 45091765.0,
      "step": 25995
    },
    {
      "entropy": 5.695248031616211,
      "epoch": 2.022874927056993,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045866006608181456,
      "loss": 4.9451,
      "mean_token_accuracy": 0.20001281648874283,
      "num_tokens": 45100206.0,
      "step": 26000
    },
    {
      "entropy": 5.664437198638916,
      "epoch": 2.0232639564287105,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004586440540901403,
      "loss": 4.9447,
      "mean_token_accuracy": 0.19852591305971146,
      "num_tokens": 45108308.0,
      "step": 26005
    },
    {
      "entropy": 5.664496278762817,
      "epoch": 2.0236529858004277,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045862803931195976,
      "loss": 5.005,
      "mean_token_accuracy": 0.20805164575576782,
      "num_tokens": 45116473.0,
      "step": 26010
    },
    {
      "entropy": 5.652438306808472,
      "epoch": 2.0240420151721454,
      "grad_norm": 1.125,
      "learning_rate": 0.0004586120217475161,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20480631589889525,
      "num_tokens": 45125913.0,
      "step": 26015
    },
    {
      "entropy": 5.667547702789307,
      "epoch": 2.024431044543863,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004585960013970522,
      "loss": 4.9369,
      "mean_token_accuracy": 0.20371063500642778,
      "num_tokens": 45134406.0,
      "step": 26020
    },
    {
      "entropy": 5.657557487487793,
      "epoch": 2.024820073915581,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004585799782608112,
      "loss": 4.9774,
      "mean_token_accuracy": 0.19581820964813232,
      "num_tokens": 45142736.0,
      "step": 26025
    },
    {
      "entropy": 5.707679748535156,
      "epoch": 2.025209103287298,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045856395233903624,
      "loss": 5.0072,
      "mean_token_accuracy": 0.20186431109905242,
      "num_tokens": 45151148.0,
      "step": 26030
    },
    {
      "entropy": 5.657434320449829,
      "epoch": 2.025598132659016,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004585479236319705,
      "loss": 4.889,
      "mean_token_accuracy": 0.21474183052778245,
      "num_tokens": 45160091.0,
      "step": 26035
    },
    {
      "entropy": 5.701471281051636,
      "epoch": 2.0259871620307335,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045853189213985714,
      "loss": 4.9933,
      "mean_token_accuracy": 0.2056482419371605,
      "num_tokens": 45168796.0,
      "step": 26040
    },
    {
      "entropy": 5.664964628219605,
      "epoch": 2.0263761914024507,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045851585786293943,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2096150740981102,
      "num_tokens": 45176891.0,
      "step": 26045
    },
    {
      "entropy": 5.690048551559448,
      "epoch": 2.0267652207741684,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045849982080146067,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20593670010566711,
      "num_tokens": 45184768.0,
      "step": 26050
    },
    {
      "entropy": 5.716774034500122,
      "epoch": 2.027154250145886,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004584837809556642,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2018101304769516,
      "num_tokens": 45193554.0,
      "step": 26055
    },
    {
      "entropy": 5.664360094070434,
      "epoch": 2.0275432795176034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045846773832579346,
      "loss": 5.0133,
      "mean_token_accuracy": 0.202696593105793,
      "num_tokens": 45202517.0,
      "step": 26060
    },
    {
      "entropy": 5.663111877441406,
      "epoch": 2.027932308889321,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045845169291209176,
      "loss": 4.9624,
      "mean_token_accuracy": 0.20017292499542236,
      "num_tokens": 45211260.0,
      "step": 26065
    },
    {
      "entropy": 5.657355260848999,
      "epoch": 2.0283213382610388,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045843564471480263,
      "loss": 5.0208,
      "mean_token_accuracy": 0.19684555232524872,
      "num_tokens": 45219828.0,
      "step": 26070
    },
    {
      "entropy": 5.755629396438598,
      "epoch": 2.0287103676327565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004584195937341695,
      "loss": 5.0255,
      "mean_token_accuracy": 0.20143416672945022,
      "num_tokens": 45228977.0,
      "step": 26075
    },
    {
      "entropy": 5.705356311798096,
      "epoch": 2.0290993970044737,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045840353997043606,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21109451055526735,
      "num_tokens": 45237191.0,
      "step": 26080
    },
    {
      "entropy": 5.6403992652893065,
      "epoch": 2.0294884263761914,
      "grad_norm": 1.25,
      "learning_rate": 0.0004583874834238458,
      "loss": 4.8953,
      "mean_token_accuracy": 0.21737575381994248,
      "num_tokens": 45245201.0,
      "step": 26085
    },
    {
      "entropy": 5.710669803619385,
      "epoch": 2.029877455747909,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004583714240946423,
      "loss": 5.0696,
      "mean_token_accuracy": 0.19903751313686371,
      "num_tokens": 45253795.0,
      "step": 26090
    },
    {
      "entropy": 5.746647024154663,
      "epoch": 2.0302664851196264,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045835536198306936,
      "loss": 5.1318,
      "mean_token_accuracy": 0.18828101456165314,
      "num_tokens": 45263403.0,
      "step": 26095
    },
    {
      "entropy": 5.708285522460938,
      "epoch": 2.030655514491344,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045833929708937067,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21156222075223924,
      "num_tokens": 45272085.0,
      "step": 26100
    },
    {
      "entropy": 5.675619077682495,
      "epoch": 2.0310445438630618,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004583232294137899,
      "loss": 4.955,
      "mean_token_accuracy": 0.20998091399669647,
      "num_tokens": 45281004.0,
      "step": 26105
    },
    {
      "entropy": 5.6479442596435545,
      "epoch": 2.031433573234779,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004583071589565709,
      "loss": 4.919,
      "mean_token_accuracy": 0.2027861073613167,
      "num_tokens": 45289950.0,
      "step": 26110
    },
    {
      "entropy": 5.661373281478882,
      "epoch": 2.0318226026064967,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004582910857179576,
      "loss": 5.0356,
      "mean_token_accuracy": 0.1914895713329315,
      "num_tokens": 45298155.0,
      "step": 26115
    },
    {
      "entropy": 5.683200836181641,
      "epoch": 2.0322116319782144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004582750096981938,
      "loss": 5.0541,
      "mean_token_accuracy": 0.20084716826677323,
      "num_tokens": 45306879.0,
      "step": 26120
    },
    {
      "entropy": 5.702759647369385,
      "epoch": 2.032600661349932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004582589308975234,
      "loss": 5.0665,
      "mean_token_accuracy": 0.19844530522823334,
      "num_tokens": 45315403.0,
      "step": 26125
    },
    {
      "entropy": 5.6728955745697025,
      "epoch": 2.0329896907216494,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045824284931619044,
      "loss": 4.9103,
      "mean_token_accuracy": 0.20828621536493303,
      "num_tokens": 45324351.0,
      "step": 26130
    },
    {
      "entropy": 5.682542514801026,
      "epoch": 2.033378720093367,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045822676495443893,
      "loss": 4.9976,
      "mean_token_accuracy": 0.20446514785289766,
      "num_tokens": 45332513.0,
      "step": 26135
    },
    {
      "entropy": 5.697757339477539,
      "epoch": 2.0337677494650848,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045821067781251284,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20063676983118056,
      "num_tokens": 45341002.0,
      "step": 26140
    },
    {
      "entropy": 5.788278579711914,
      "epoch": 2.034156778836802,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045819458789065633,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20564684569835662,
      "num_tokens": 45349650.0,
      "step": 26145
    },
    {
      "entropy": 5.760427618026734,
      "epoch": 2.0345458082085197,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004581784951891135,
      "loss": 5.1386,
      "mean_token_accuracy": 0.1907847613096237,
      "num_tokens": 45359189.0,
      "step": 26150
    },
    {
      "entropy": 5.63937201499939,
      "epoch": 2.0349348375802374,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004581623997081286,
      "loss": 4.9085,
      "mean_token_accuracy": 0.20986852347850798,
      "num_tokens": 45367274.0,
      "step": 26155
    },
    {
      "entropy": 5.646141910552979,
      "epoch": 2.0353238669519547,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004581463014479459,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21152779906988145,
      "num_tokens": 45375424.0,
      "step": 26160
    },
    {
      "entropy": 5.720108318328857,
      "epoch": 2.0357128963236724,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004581302004088095,
      "loss": 5.0009,
      "mean_token_accuracy": 0.20606516748666764,
      "num_tokens": 45384333.0,
      "step": 26165
    },
    {
      "entropy": 5.716523027420044,
      "epoch": 2.03610192569539,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004581140965909638,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19511331170797347,
      "num_tokens": 45393309.0,
      "step": 26170
    },
    {
      "entropy": 5.718493175506592,
      "epoch": 2.0364909550671078,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004580979899946531,
      "loss": 5.0598,
      "mean_token_accuracy": 0.196629536151886,
      "num_tokens": 45401590.0,
      "step": 26175
    },
    {
      "entropy": 5.700514316558838,
      "epoch": 2.036879984438825,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004580818806201219,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20037599503993989,
      "num_tokens": 45409458.0,
      "step": 26180
    },
    {
      "entropy": 5.670484495162964,
      "epoch": 2.0372690138105427,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045806576846761453,
      "loss": 4.935,
      "mean_token_accuracy": 0.20506274700164795,
      "num_tokens": 45417777.0,
      "step": 26185
    },
    {
      "entropy": 5.708735132217408,
      "epoch": 2.0376580431822604,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045804965353737556,
      "loss": 5.027,
      "mean_token_accuracy": 0.20171735137701036,
      "num_tokens": 45425489.0,
      "step": 26190
    },
    {
      "entropy": 5.635439920425415,
      "epoch": 2.0380470725539777,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004580335358296494,
      "loss": 4.9504,
      "mean_token_accuracy": 0.20681928247213363,
      "num_tokens": 45433770.0,
      "step": 26195
    },
    {
      "entropy": 5.712325191497802,
      "epoch": 2.0384361019256954,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045801741534468065,
      "loss": 5.025,
      "mean_token_accuracy": 0.20346264988183976,
      "num_tokens": 45442607.0,
      "step": 26200
    },
    {
      "entropy": 5.673827123641968,
      "epoch": 2.038825131297413,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004580012920827139,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20656422525644302,
      "num_tokens": 45451671.0,
      "step": 26205
    },
    {
      "entropy": 5.714946651458741,
      "epoch": 2.0392141606691303,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004579851660439939,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2069881707429886,
      "num_tokens": 45459861.0,
      "step": 26210
    },
    {
      "entropy": 5.6356336116790775,
      "epoch": 2.039603190040848,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045796903722876523,
      "loss": 4.8872,
      "mean_token_accuracy": 0.21211285144090652,
      "num_tokens": 45468623.0,
      "step": 26215
    },
    {
      "entropy": 5.641687107086182,
      "epoch": 2.0399922194125657,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004579529056372726,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2033635601401329,
      "num_tokens": 45476763.0,
      "step": 26220
    },
    {
      "entropy": 5.622302103042602,
      "epoch": 2.0403812487842834,
      "grad_norm": 1.25,
      "learning_rate": 0.0004579367712697609,
      "loss": 4.9474,
      "mean_token_accuracy": 0.2062364012002945,
      "num_tokens": 45485069.0,
      "step": 26225
    },
    {
      "entropy": 5.610862350463867,
      "epoch": 2.0407702781560006,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045792063412647475,
      "loss": 5.0762,
      "mean_token_accuracy": 0.1949963241815567,
      "num_tokens": 45494236.0,
      "step": 26230
    },
    {
      "entropy": 5.640335321426392,
      "epoch": 2.0411593075277183,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045790449420765925,
      "loss": 4.9463,
      "mean_token_accuracy": 0.19776529371738433,
      "num_tokens": 45503257.0,
      "step": 26235
    },
    {
      "entropy": 5.7358770847320555,
      "epoch": 2.041548336899436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045788835151355914,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20274417996406555,
      "num_tokens": 45511544.0,
      "step": 26240
    },
    {
      "entropy": 5.670800971984863,
      "epoch": 2.0419373662711533,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045787220604441933,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20595357120037078,
      "num_tokens": 45519951.0,
      "step": 26245
    },
    {
      "entropy": 5.740437793731689,
      "epoch": 2.042326395642871,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045785605780048497,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1989247754216194,
      "num_tokens": 45527978.0,
      "step": 26250
    },
    {
      "entropy": 5.748931264877319,
      "epoch": 2.0427154250145887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045783990678200086,
      "loss": 5.026,
      "mean_token_accuracy": 0.2030486062169075,
      "num_tokens": 45537103.0,
      "step": 26255
    },
    {
      "entropy": 5.679254245758057,
      "epoch": 2.043104454386306,
      "grad_norm": 1.25,
      "learning_rate": 0.00045782375298921227,
      "loss": 4.9534,
      "mean_token_accuracy": 0.20283795446157454,
      "num_tokens": 45545136.0,
      "step": 26260
    },
    {
      "entropy": 5.757080316543579,
      "epoch": 2.0434934837580236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004578075964223642,
      "loss": 5.0862,
      "mean_token_accuracy": 0.1963065430521965,
      "num_tokens": 45553709.0,
      "step": 26265
    },
    {
      "entropy": 5.62369065284729,
      "epoch": 2.0438825131297413,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004577914370817018,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20451295226812363,
      "num_tokens": 45562947.0,
      "step": 26270
    },
    {
      "entropy": 5.649138450622559,
      "epoch": 2.044271542501459,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045777527496747034,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2002468451857567,
      "num_tokens": 45571319.0,
      "step": 26275
    },
    {
      "entropy": 5.683254051208496,
      "epoch": 2.0446605718731763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045775911007991493,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20405613332986833,
      "num_tokens": 45580071.0,
      "step": 26280
    },
    {
      "entropy": 5.636892318725586,
      "epoch": 2.045049601244894,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045774294241928093,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21173634976148606,
      "num_tokens": 45587998.0,
      "step": 26285
    },
    {
      "entropy": 5.691944932937622,
      "epoch": 2.0454386306166117,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004577267719858137,
      "loss": 4.9876,
      "mean_token_accuracy": 0.21124788969755173,
      "num_tokens": 45597109.0,
      "step": 26290
    },
    {
      "entropy": 5.70725564956665,
      "epoch": 2.045827659988329,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045771059877975853,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20311434119939803,
      "num_tokens": 45605699.0,
      "step": 26295
    },
    {
      "entropy": 5.789184284210205,
      "epoch": 2.0462166893600466,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004576944228013608,
      "loss": 5.1307,
      "mean_token_accuracy": 0.1929370179772377,
      "num_tokens": 45613634.0,
      "step": 26300
    },
    {
      "entropy": 5.68727388381958,
      "epoch": 2.0466057187317643,
      "grad_norm": 1.15625,
      "learning_rate": 0.000457678244050866,
      "loss": 4.8788,
      "mean_token_accuracy": 0.21207094490528106,
      "num_tokens": 45621814.0,
      "step": 26305
    },
    {
      "entropy": 5.725837469100952,
      "epoch": 2.046994748103482,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004576620625285196,
      "loss": 5.0985,
      "mean_token_accuracy": 0.1956136018037796,
      "num_tokens": 45629386.0,
      "step": 26310
    },
    {
      "entropy": 5.579510450363159,
      "epoch": 2.0473837774751993,
      "grad_norm": 1.375,
      "learning_rate": 0.00045764587823456717,
      "loss": 4.8863,
      "mean_token_accuracy": 0.210037724673748,
      "num_tokens": 45637565.0,
      "step": 26315
    },
    {
      "entropy": 5.66849536895752,
      "epoch": 2.047772806846917,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045762969116925424,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20522323846817017,
      "num_tokens": 45646876.0,
      "step": 26320
    },
    {
      "entropy": 5.656621408462525,
      "epoch": 2.0481618362186347,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045761350133282643,
      "loss": 4.8574,
      "mean_token_accuracy": 0.20942877531051635,
      "num_tokens": 45655445.0,
      "step": 26325
    },
    {
      "entropy": 5.615228605270386,
      "epoch": 2.048550865590352,
      "grad_norm": 1.25,
      "learning_rate": 0.00045759730872552937,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2006406456232071,
      "num_tokens": 45663908.0,
      "step": 26330
    },
    {
      "entropy": 5.703623247146607,
      "epoch": 2.0489398949620696,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004575811133476088,
      "loss": 5.0161,
      "mean_token_accuracy": 0.20869170874357224,
      "num_tokens": 45672842.0,
      "step": 26335
    },
    {
      "entropy": 5.691976737976074,
      "epoch": 2.0493289243337873,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045756491519931047,
      "loss": 4.9823,
      "mean_token_accuracy": 0.20560858845710756,
      "num_tokens": 45680988.0,
      "step": 26340
    },
    {
      "entropy": 5.679571151733398,
      "epoch": 2.0497179537055046,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004575487142808801,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20587874054908753,
      "num_tokens": 45690165.0,
      "step": 26345
    },
    {
      "entropy": 5.702537584304809,
      "epoch": 2.0501069830772223,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004575325105925636,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20308963805437089,
      "num_tokens": 45698721.0,
      "step": 26350
    },
    {
      "entropy": 5.687490701675415,
      "epoch": 2.05049601244894,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045751630413460665,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2050777330994606,
      "num_tokens": 45707374.0,
      "step": 26355
    },
    {
      "entropy": 5.64647946357727,
      "epoch": 2.0508850418206577,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004575000949072554,
      "loss": 4.8736,
      "mean_token_accuracy": 0.21584661453962325,
      "num_tokens": 45715756.0,
      "step": 26360
    },
    {
      "entropy": 5.620793962478638,
      "epoch": 2.051274071192375,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004574838829107557,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2119452565908432,
      "num_tokens": 45724173.0,
      "step": 26365
    },
    {
      "entropy": 5.708711242675781,
      "epoch": 2.0516631005640926,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004574676681453535,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2051585391163826,
      "num_tokens": 45732330.0,
      "step": 26370
    },
    {
      "entropy": 5.683817195892334,
      "epoch": 2.0520521299358103,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045745145061129485,
      "loss": 5.003,
      "mean_token_accuracy": 0.20448270589113235,
      "num_tokens": 45741489.0,
      "step": 26375
    },
    {
      "entropy": 5.584952688217163,
      "epoch": 2.0524411593075276,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045743523030882584,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2121899351477623,
      "num_tokens": 45750816.0,
      "step": 26380
    },
    {
      "entropy": 5.746816349029541,
      "epoch": 2.0528301886792453,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004574190072381926,
      "loss": 5.0839,
      "mean_token_accuracy": 0.2041749283671379,
      "num_tokens": 45759393.0,
      "step": 26385
    },
    {
      "entropy": 5.731736278533935,
      "epoch": 2.053219218050963,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004574027813996413,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20231543183326722,
      "num_tokens": 45767749.0,
      "step": 26390
    },
    {
      "entropy": 5.668630790710449,
      "epoch": 2.05360824742268,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004573865527934181,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20133038759231567,
      "num_tokens": 45776455.0,
      "step": 26395
    },
    {
      "entropy": 5.6749145030975345,
      "epoch": 2.053997276794398,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004573703214197692,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2101456731557846,
      "num_tokens": 45784869.0,
      "step": 26400
    },
    {
      "entropy": 5.655792617797852,
      "epoch": 2.0543863061661156,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045735408727894095,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2096520632505417,
      "num_tokens": 45793956.0,
      "step": 26405
    },
    {
      "entropy": 5.686174297332764,
      "epoch": 2.0547753355378333,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045733785037117977,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20019271671772004,
      "num_tokens": 45801918.0,
      "step": 26410
    },
    {
      "entropy": 5.684372043609619,
      "epoch": 2.0551643649095506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004573216106967319,
      "loss": 4.9606,
      "mean_token_accuracy": 0.20564018040895463,
      "num_tokens": 45810026.0,
      "step": 26415
    },
    {
      "entropy": 5.613135528564453,
      "epoch": 2.0555533942812683,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045730536825584365,
      "loss": 4.9179,
      "mean_token_accuracy": 0.20611971467733384,
      "num_tokens": 45818101.0,
      "step": 26420
    },
    {
      "entropy": 5.688903331756592,
      "epoch": 2.055942423652986,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045728912304876176,
      "loss": 5.0022,
      "mean_token_accuracy": 0.2040456250309944,
      "num_tokens": 45826467.0,
      "step": 26425
    },
    {
      "entropy": 5.69919638633728,
      "epoch": 2.056331453024703,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004572728750757326,
      "loss": 4.9572,
      "mean_token_accuracy": 0.20351143926382065,
      "num_tokens": 45835329.0,
      "step": 26430
    },
    {
      "entropy": 5.708038568496704,
      "epoch": 2.056720482396421,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004572566243370025,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20818026065826417,
      "num_tokens": 45844449.0,
      "step": 26435
    },
    {
      "entropy": 5.634861755371094,
      "epoch": 2.0571095117681386,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004572403708328183,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20137871205806732,
      "num_tokens": 45854042.0,
      "step": 26440
    },
    {
      "entropy": 5.681427764892578,
      "epoch": 2.057498541139856,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004572241145634266,
      "loss": 4.9814,
      "mean_token_accuracy": 0.20955270975828172,
      "num_tokens": 45863164.0,
      "step": 26445
    },
    {
      "entropy": 5.715148401260376,
      "epoch": 2.0578875705115736,
      "grad_norm": 1.109375,
      "learning_rate": 0.000457207855529074,
      "loss": 5.0419,
      "mean_token_accuracy": 0.20494148433208464,
      "num_tokens": 45872054.0,
      "step": 26450
    },
    {
      "entropy": 5.668500757217407,
      "epoch": 2.0582765998832913,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045719159373000713,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20809558480978013,
      "num_tokens": 45880887.0,
      "step": 26455
    },
    {
      "entropy": 5.775046920776367,
      "epoch": 2.058665629255009,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004571753291664729,
      "loss": 5.101,
      "mean_token_accuracy": 0.19785018116235734,
      "num_tokens": 45890078.0,
      "step": 26460
    },
    {
      "entropy": 5.705823087692261,
      "epoch": 2.059054658626726,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004571590618387179,
      "loss": 5.0306,
      "mean_token_accuracy": 0.2062179207801819,
      "num_tokens": 45899803.0,
      "step": 26465
    },
    {
      "entropy": 5.766442775726318,
      "epoch": 2.059443687998444,
      "grad_norm": 1.25,
      "learning_rate": 0.0004571427917469892,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20223882049322128,
      "num_tokens": 45908851.0,
      "step": 26470
    },
    {
      "entropy": 5.6849054336547855,
      "epoch": 2.0598327173701616,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045712651889153345,
      "loss": 4.9733,
      "mean_token_accuracy": 0.2072332426905632,
      "num_tokens": 45917891.0,
      "step": 26475
    },
    {
      "entropy": 5.693885135650635,
      "epoch": 2.060221746741879,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045711024327259764,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20345047414302825,
      "num_tokens": 45926185.0,
      "step": 26480
    },
    {
      "entropy": 5.6979045391082765,
      "epoch": 2.0606107761135966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045709396489042884,
      "loss": 5.0827,
      "mean_token_accuracy": 0.1995086655020714,
      "num_tokens": 45935064.0,
      "step": 26485
    },
    {
      "entropy": 5.6708272933959964,
      "epoch": 2.0609998054853143,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045707768374527385,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20018247067928313,
      "num_tokens": 45943216.0,
      "step": 26490
    },
    {
      "entropy": 5.693432903289795,
      "epoch": 2.0613888348570315,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004570613998373799,
      "loss": 5.0832,
      "mean_token_accuracy": 0.19901855140924454,
      "num_tokens": 45952178.0,
      "step": 26495
    },
    {
      "entropy": 5.765191698074341,
      "epoch": 2.061777864228749,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045704511316699395,
      "loss": 5.0859,
      "mean_token_accuracy": 0.19375887215137483,
      "num_tokens": 45960962.0,
      "step": 26500
    },
    {
      "entropy": 5.649706411361694,
      "epoch": 2.062166893600467,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045702882373436317,
      "loss": 4.8383,
      "mean_token_accuracy": 0.21082568168640137,
      "num_tokens": 45969527.0,
      "step": 26505
    },
    {
      "entropy": 5.659423923492431,
      "epoch": 2.0625559229721846,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004570125315397347,
      "loss": 4.966,
      "mean_token_accuracy": 0.2027256652712822,
      "num_tokens": 45978258.0,
      "step": 26510
    },
    {
      "entropy": 5.553504514694214,
      "epoch": 2.062944952343902,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004569962365833558,
      "loss": 4.9224,
      "mean_token_accuracy": 0.21248174458742142,
      "num_tokens": 45986647.0,
      "step": 26515
    },
    {
      "entropy": 5.641728258132934,
      "epoch": 2.0633339817156195,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045697993886547374,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2121603637933731,
      "num_tokens": 45995458.0,
      "step": 26520
    },
    {
      "entropy": 5.806697225570678,
      "epoch": 2.0637230110873372,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045696363838633566,
      "loss": 5.0676,
      "mean_token_accuracy": 0.1999026581645012,
      "num_tokens": 46004293.0,
      "step": 26525
    },
    {
      "entropy": 5.72117772102356,
      "epoch": 2.0641120404590545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045694733514618904,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18797560632228852,
      "num_tokens": 46013752.0,
      "step": 26530
    },
    {
      "entropy": 5.69691686630249,
      "epoch": 2.064501069830772,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004569310291452812,
      "loss": 5.0505,
      "mean_token_accuracy": 0.19965075850486755,
      "num_tokens": 46021935.0,
      "step": 26535
    },
    {
      "entropy": 5.599312734603882,
      "epoch": 2.06489009920249,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004569147203838595,
      "loss": 4.9306,
      "mean_token_accuracy": 0.21271594613790512,
      "num_tokens": 46030567.0,
      "step": 26540
    },
    {
      "entropy": 5.727325105667115,
      "epoch": 2.065279128574207,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045689840886217157,
      "loss": 5.0688,
      "mean_token_accuracy": 0.1979468584060669,
      "num_tokens": 46038764.0,
      "step": 26545
    },
    {
      "entropy": 5.599790620803833,
      "epoch": 2.065668157945925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045688209458046475,
      "loss": 4.8344,
      "mean_token_accuracy": 0.21708200126886368,
      "num_tokens": 46047393.0,
      "step": 26550
    },
    {
      "entropy": 5.652333402633667,
      "epoch": 2.0660571873176425,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045686577753898663,
      "loss": 4.958,
      "mean_token_accuracy": 0.19636675864458084,
      "num_tokens": 46055748.0,
      "step": 26555
    },
    {
      "entropy": 5.686590623855591,
      "epoch": 2.0664462166893602,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045684945773798483,
      "loss": 4.945,
      "mean_token_accuracy": 0.20345400124788285,
      "num_tokens": 46063888.0,
      "step": 26560
    },
    {
      "entropy": 5.596152019500733,
      "epoch": 2.0668352460610775,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000456833135177707,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21638067960739135,
      "num_tokens": 46073031.0,
      "step": 26565
    },
    {
      "entropy": 5.69827184677124,
      "epoch": 2.067224275432795,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004568168098584007,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20016423761844634,
      "num_tokens": 46081204.0,
      "step": 26570
    },
    {
      "entropy": 5.755049419403076,
      "epoch": 2.067613304804513,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004568004817803137,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21224044412374496,
      "num_tokens": 46089842.0,
      "step": 26575
    },
    {
      "entropy": 5.668024396896362,
      "epoch": 2.06800233417623,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004567841509436937,
      "loss": 4.9849,
      "mean_token_accuracy": 0.20969805270433425,
      "num_tokens": 46098400.0,
      "step": 26580
    },
    {
      "entropy": 5.6796348094940186,
      "epoch": 2.068391363547948,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004567678173487887,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19287682622671126,
      "num_tokens": 46106948.0,
      "step": 26585
    },
    {
      "entropy": 5.71806960105896,
      "epoch": 2.0687803929196655,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004567514809958462,
      "loss": 4.9624,
      "mean_token_accuracy": 0.20698493868112564,
      "num_tokens": 46115361.0,
      "step": 26590
    },
    {
      "entropy": 5.683470582962036,
      "epoch": 2.069169422291383,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004567351418851144,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2047705575823784,
      "num_tokens": 46123658.0,
      "step": 26595
    },
    {
      "entropy": 5.72746844291687,
      "epoch": 2.0695584516631005,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045671880001684113,
      "loss": 5.027,
      "mean_token_accuracy": 0.19979496002197267,
      "num_tokens": 46132511.0,
      "step": 26600
    },
    {
      "entropy": 5.700604629516602,
      "epoch": 2.069947481034818,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045670245539127413,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19790847450494767,
      "num_tokens": 46141337.0,
      "step": 26605
    },
    {
      "entropy": 5.719999837875366,
      "epoch": 2.070336510406536,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045668610800866173,
      "loss": 5.0203,
      "mean_token_accuracy": 0.19888515025377274,
      "num_tokens": 46149161.0,
      "step": 26610
    },
    {
      "entropy": 5.771575307846069,
      "epoch": 2.070725539778253,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045666975786925177,
      "loss": 5.0804,
      "mean_token_accuracy": 0.19175949394702912,
      "num_tokens": 46157497.0,
      "step": 26615
    },
    {
      "entropy": 5.666361331939697,
      "epoch": 2.071114569149971,
      "grad_norm": 1.25,
      "learning_rate": 0.0004566534049732923,
      "loss": 4.967,
      "mean_token_accuracy": 0.19793857634067535,
      "num_tokens": 46165989.0,
      "step": 26620
    },
    {
      "entropy": 5.7368158340454105,
      "epoch": 2.0715035985216885,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045663704932103166,
      "loss": 4.9696,
      "mean_token_accuracy": 0.20906771570444108,
      "num_tokens": 46174847.0,
      "step": 26625
    },
    {
      "entropy": 5.684810972213745,
      "epoch": 2.071892627893406,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045662069091271785,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2091636538505554,
      "num_tokens": 46183680.0,
      "step": 26630
    },
    {
      "entropy": 5.666780424118042,
      "epoch": 2.0722816572651235,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045660432974859924,
      "loss": 4.9823,
      "mean_token_accuracy": 0.20436631590127946,
      "num_tokens": 46192878.0,
      "step": 26635
    },
    {
      "entropy": 5.74148736000061,
      "epoch": 2.072670686636841,
      "grad_norm": 1.25,
      "learning_rate": 0.00045658796582892383,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19706131666898727,
      "num_tokens": 46200891.0,
      "step": 26640
    },
    {
      "entropy": 5.6107875347137455,
      "epoch": 2.073059716008559,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045657159915394013,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21221780329942702,
      "num_tokens": 46210138.0,
      "step": 26645
    },
    {
      "entropy": 5.686340618133545,
      "epoch": 2.073448745380276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004565552297238964,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20624729245901108,
      "num_tokens": 46218729.0,
      "step": 26650
    },
    {
      "entropy": 5.664607048034668,
      "epoch": 2.073837774751994,
      "grad_norm": 1.25,
      "learning_rate": 0.000456538857539041,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20771866887807847,
      "num_tokens": 46228462.0,
      "step": 26655
    },
    {
      "entropy": 5.674657201766967,
      "epoch": 2.0742268041237115,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045652248259962254,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2161584675312042,
      "num_tokens": 46237417.0,
      "step": 26660
    },
    {
      "entropy": 5.668230676651001,
      "epoch": 2.0746158334954288,
      "grad_norm": 1.25,
      "learning_rate": 0.0004565061049058892,
      "loss": 5.0293,
      "mean_token_accuracy": 0.19325605630874634,
      "num_tokens": 46246344.0,
      "step": 26665
    },
    {
      "entropy": 5.711876106262207,
      "epoch": 2.0750048628671465,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045648972445808963,
      "loss": 5.0573,
      "mean_token_accuracy": 0.19740857928991318,
      "num_tokens": 46255286.0,
      "step": 26670
    },
    {
      "entropy": 5.7000274658203125,
      "epoch": 2.075393892238864,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045647334125647235,
      "loss": 5.0123,
      "mean_token_accuracy": 0.20596438646316528,
      "num_tokens": 46264737.0,
      "step": 26675
    },
    {
      "entropy": 5.66577296257019,
      "epoch": 2.0757829216105814,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000456456955301286,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2120671570301056,
      "num_tokens": 46273159.0,
      "step": 26680
    },
    {
      "entropy": 5.737722635269165,
      "epoch": 2.076171950982299,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004564405665927791,
      "loss": 5.0123,
      "mean_token_accuracy": 0.19954560995101928,
      "num_tokens": 46281559.0,
      "step": 26685
    },
    {
      "entropy": 5.630560398101807,
      "epoch": 2.076560980354017,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045642417513120043,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21172236651182175,
      "num_tokens": 46289941.0,
      "step": 26690
    },
    {
      "entropy": 5.608247089385986,
      "epoch": 2.076950009725734,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045640778091679876,
      "loss": 4.9475,
      "mean_token_accuracy": 0.20516565144062043,
      "num_tokens": 46298711.0,
      "step": 26695
    },
    {
      "entropy": 5.679301929473877,
      "epoch": 2.0773390390974518,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004563913839498226,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21014782637357712,
      "num_tokens": 46306956.0,
      "step": 26700
    },
    {
      "entropy": 5.698627614974976,
      "epoch": 2.0777280684691695,
      "grad_norm": 1.25,
      "learning_rate": 0.000456374984230521,
      "loss": 4.9489,
      "mean_token_accuracy": 0.205281962454319,
      "num_tokens": 46315165.0,
      "step": 26705
    },
    {
      "entropy": 5.6624030590057375,
      "epoch": 2.078117097840887,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004563585817591427,
      "loss": 4.8634,
      "mean_token_accuracy": 0.217240047454834,
      "num_tokens": 46323347.0,
      "step": 26710
    },
    {
      "entropy": 5.705899095535278,
      "epoch": 2.0785061272126044,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004563421765359365,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20630285888910294,
      "num_tokens": 46331302.0,
      "step": 26715
    },
    {
      "entropy": 5.705237102508545,
      "epoch": 2.078895156584322,
      "grad_norm": 1.25,
      "learning_rate": 0.00045632576856115156,
      "loss": 5.015,
      "mean_token_accuracy": 0.19834015220403672,
      "num_tokens": 46339050.0,
      "step": 26720
    },
    {
      "entropy": 5.686740970611572,
      "epoch": 2.07928418595604,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045630935783503657,
      "loss": 4.9927,
      "mean_token_accuracy": 0.20858113914728166,
      "num_tokens": 46347350.0,
      "step": 26725
    },
    {
      "entropy": 5.7930632591247555,
      "epoch": 2.079673215327757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004562929443578407,
      "loss": 5.0447,
      "mean_token_accuracy": 0.19489240944385527,
      "num_tokens": 46356951.0,
      "step": 26730
    },
    {
      "entropy": 5.786788654327393,
      "epoch": 2.0800622446994748,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004562765281298129,
      "loss": 5.0744,
      "mean_token_accuracy": 0.2056357353925705,
      "num_tokens": 46367056.0,
      "step": 26735
    },
    {
      "entropy": 5.683865022659302,
      "epoch": 2.0804512740711925,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004562601091512024,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20761232525110246,
      "num_tokens": 46375668.0,
      "step": 26740
    },
    {
      "entropy": 5.637686157226563,
      "epoch": 2.08084030344291,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004562436874222582,
      "loss": 4.9659,
      "mean_token_accuracy": 0.20612082183361052,
      "num_tokens": 46384460.0,
      "step": 26745
    },
    {
      "entropy": 5.565558624267578,
      "epoch": 2.0812293328146274,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045622726294322955,
      "loss": 4.8151,
      "mean_token_accuracy": 0.21240213960409166,
      "num_tokens": 46393349.0,
      "step": 26750
    },
    {
      "entropy": 5.714799880981445,
      "epoch": 2.081618362186345,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045621083571436563,
      "loss": 4.9631,
      "mean_token_accuracy": 0.20710004419088363,
      "num_tokens": 46403081.0,
      "step": 26755
    },
    {
      "entropy": 5.66352653503418,
      "epoch": 2.082007391558063,
      "grad_norm": 1.125,
      "learning_rate": 0.0004561944057359157,
      "loss": 4.8824,
      "mean_token_accuracy": 0.21396737694740295,
      "num_tokens": 46412219.0,
      "step": 26760
    },
    {
      "entropy": 5.704862451553344,
      "epoch": 2.08239642092978,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004561779730081291,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20653637498617172,
      "num_tokens": 46420811.0,
      "step": 26765
    },
    {
      "entropy": 5.6315137386322025,
      "epoch": 2.0827854503014978,
      "grad_norm": 1.125,
      "learning_rate": 0.0004561615375312551,
      "loss": 4.9319,
      "mean_token_accuracy": 0.20853445827960967,
      "num_tokens": 46429431.0,
      "step": 26770
    },
    {
      "entropy": 5.685202503204346,
      "epoch": 2.0831744796732155,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045614509930554304,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20030997544527054,
      "num_tokens": 46438344.0,
      "step": 26775
    },
    {
      "entropy": 5.7133910179138185,
      "epoch": 2.0835635090449327,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045612865833124253,
      "loss": 5.0412,
      "mean_token_accuracy": 0.20069828480482102,
      "num_tokens": 46446960.0,
      "step": 26780
    },
    {
      "entropy": 5.667187738418579,
      "epoch": 2.0839525384166504,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004561122146086029,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20699167400598525,
      "num_tokens": 46454862.0,
      "step": 26785
    },
    {
      "entropy": 5.710286474227905,
      "epoch": 2.084341567788368,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004560957681378737,
      "loss": 5.058,
      "mean_token_accuracy": 0.20312601029872895,
      "num_tokens": 46463208.0,
      "step": 26790
    },
    {
      "entropy": 5.750453662872315,
      "epoch": 2.084730597160086,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045607931891930445,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19883539974689485,
      "num_tokens": 46472253.0,
      "step": 26795
    },
    {
      "entropy": 5.800017404556274,
      "epoch": 2.085119626531803,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004560628669531447,
      "loss": 5.0839,
      "mean_token_accuracy": 0.19506797939538956,
      "num_tokens": 46482133.0,
      "step": 26800
    },
    {
      "entropy": 5.707293558120727,
      "epoch": 2.0855086559035207,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045604641223964416,
      "loss": 5.0334,
      "mean_token_accuracy": 0.2041900634765625,
      "num_tokens": 46491566.0,
      "step": 26805
    },
    {
      "entropy": 5.692251348495484,
      "epoch": 2.0858976852752384,
      "grad_norm": 1.125,
      "learning_rate": 0.00045602995477905247,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19303447753190994,
      "num_tokens": 46501283.0,
      "step": 26810
    },
    {
      "entropy": 5.754811334609985,
      "epoch": 2.0862867146469557,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004560134945716194,
      "loss": 4.984,
      "mean_token_accuracy": 0.20727898478507994,
      "num_tokens": 46510384.0,
      "step": 26815
    },
    {
      "entropy": 5.765845680236817,
      "epoch": 2.0866757440186734,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045599703161759464,
      "loss": 4.986,
      "mean_token_accuracy": 0.20449990332126616,
      "num_tokens": 46519202.0,
      "step": 26820
    },
    {
      "entropy": 5.6790642738342285,
      "epoch": 2.087064773390391,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045598056591722805,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2137366786599159,
      "num_tokens": 46528030.0,
      "step": 26825
    },
    {
      "entropy": 5.688123512268066,
      "epoch": 2.0874538027621083,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045596409747076936,
      "loss": 5.0282,
      "mean_token_accuracy": 0.2005438193678856,
      "num_tokens": 46537222.0,
      "step": 26830
    },
    {
      "entropy": 5.7417432308197025,
      "epoch": 2.087842832133826,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004559476262784686,
      "loss": 4.9632,
      "mean_token_accuracy": 0.20855283439159394,
      "num_tokens": 46546486.0,
      "step": 26835
    },
    {
      "entropy": 5.739138650894165,
      "epoch": 2.0882318615055437,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004559311523405755,
      "loss": 5.0072,
      "mean_token_accuracy": 0.20598165243864058,
      "num_tokens": 46554518.0,
      "step": 26840
    },
    {
      "entropy": 5.704701042175293,
      "epoch": 2.0886208908772614,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004559146756573402,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20343564748764037,
      "num_tokens": 46563183.0,
      "step": 26845
    },
    {
      "entropy": 5.6879621028900145,
      "epoch": 2.0890099202489787,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045589819622901274,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21389485597610475,
      "num_tokens": 46571514.0,
      "step": 26850
    },
    {
      "entropy": 5.712914991378784,
      "epoch": 2.0893989496206964,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000455881714055843,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2107968047261238,
      "num_tokens": 46580827.0,
      "step": 26855
    },
    {
      "entropy": 5.689039421081543,
      "epoch": 2.089787978992414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045586522913808114,
      "loss": 5.0644,
      "mean_token_accuracy": 0.19669135808944702,
      "num_tokens": 46591254.0,
      "step": 26860
    },
    {
      "entropy": 5.7317054271698,
      "epoch": 2.0901770083641313,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004558487414759773,
      "loss": 4.935,
      "mean_token_accuracy": 0.20484873056411743,
      "num_tokens": 46599576.0,
      "step": 26865
    },
    {
      "entropy": 5.698754119873047,
      "epoch": 2.090566037735849,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045583225106978175,
      "loss": 5.0236,
      "mean_token_accuracy": 0.2017316475510597,
      "num_tokens": 46607607.0,
      "step": 26870
    },
    {
      "entropy": 5.584449195861817,
      "epoch": 2.0909550671075667,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004558157579197446,
      "loss": 4.9382,
      "mean_token_accuracy": 0.20413241386413575,
      "num_tokens": 46616446.0,
      "step": 26875
    },
    {
      "entropy": 5.637922716140747,
      "epoch": 2.091344096479284,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045579926202611603,
      "loss": 4.9386,
      "mean_token_accuracy": 0.20444327890872954,
      "num_tokens": 46624875.0,
      "step": 26880
    },
    {
      "entropy": 5.6667030334472654,
      "epoch": 2.0917331258510017,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004557827633891465,
      "loss": 4.9226,
      "mean_token_accuracy": 0.20650898218154906,
      "num_tokens": 46633354.0,
      "step": 26885
    },
    {
      "entropy": 5.724954605102539,
      "epoch": 2.0921221552227194,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004557662620090863,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20998506098985673,
      "num_tokens": 46642034.0,
      "step": 26890
    },
    {
      "entropy": 5.678207683563232,
      "epoch": 2.092511184594437,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045574975788618564,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20987818241119385,
      "num_tokens": 46650675.0,
      "step": 26895
    },
    {
      "entropy": 5.656865644454956,
      "epoch": 2.0929002139661543,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004557332510206953,
      "loss": 4.9141,
      "mean_token_accuracy": 0.20876990556716918,
      "num_tokens": 46658834.0,
      "step": 26900
    },
    {
      "entropy": 5.6131758213043215,
      "epoch": 2.093289243337872,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004557167414128654,
      "loss": 4.8736,
      "mean_token_accuracy": 0.21493054032325745,
      "num_tokens": 46667386.0,
      "step": 26905
    },
    {
      "entropy": 5.634513473510742,
      "epoch": 2.0936782727095897,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004557002290629467,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21589386016130446,
      "num_tokens": 46675492.0,
      "step": 26910
    },
    {
      "entropy": 5.641925716400147,
      "epoch": 2.094067302081307,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045568371397118954,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20418912321329116,
      "num_tokens": 46684079.0,
      "step": 26915
    },
    {
      "entropy": 5.662522649765014,
      "epoch": 2.0944563314530247,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004556671961378446,
      "loss": 5.016,
      "mean_token_accuracy": 0.2012130632996559,
      "num_tokens": 46693059.0,
      "step": 26920
    },
    {
      "entropy": 5.732287311553955,
      "epoch": 2.0948453608247424,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045565067556316264,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2093151718378067,
      "num_tokens": 46701639.0,
      "step": 26925
    },
    {
      "entropy": 5.671289300918579,
      "epoch": 2.0952343901964596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004556341522473941,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20570600926876068,
      "num_tokens": 46710870.0,
      "step": 26930
    },
    {
      "entropy": 5.731515979766845,
      "epoch": 2.0956234195681773,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004556176261907899,
      "loss": 5.0357,
      "mean_token_accuracy": 0.1991974487900734,
      "num_tokens": 46719236.0,
      "step": 26935
    },
    {
      "entropy": 5.706711912155152,
      "epoch": 2.096012448939895,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004556010973936006,
      "loss": 4.9814,
      "mean_token_accuracy": 0.20421761125326157,
      "num_tokens": 46727595.0,
      "step": 26940
    },
    {
      "entropy": 5.700484085083008,
      "epoch": 2.0964014783116127,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004555845658560772,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2060418799519539,
      "num_tokens": 46735874.0,
      "step": 26945
    },
    {
      "entropy": 5.639110946655274,
      "epoch": 2.09679050768333,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045556803157847036,
      "loss": 4.9927,
      "mean_token_accuracy": 0.20640319138765334,
      "num_tokens": 46744738.0,
      "step": 26950
    },
    {
      "entropy": 5.652397632598877,
      "epoch": 2.0971795370550477,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045555149456103113,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2048835426568985,
      "num_tokens": 46753252.0,
      "step": 26955
    },
    {
      "entropy": 5.669762134552002,
      "epoch": 2.0975685664267654,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004555349548040102,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21061916947364806,
      "num_tokens": 46761234.0,
      "step": 26960
    },
    {
      "entropy": 5.681529474258423,
      "epoch": 2.0979575957984826,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004555184123076589,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19017570465803146,
      "num_tokens": 46770087.0,
      "step": 26965
    },
    {
      "entropy": 5.578570175170898,
      "epoch": 2.0983466251702003,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045550186707222785,
      "loss": 4.9257,
      "mean_token_accuracy": 0.21004963368177415,
      "num_tokens": 46778985.0,
      "step": 26970
    },
    {
      "entropy": 5.693442153930664,
      "epoch": 2.098735654541918,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004554853190979683,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20307573825120925,
      "num_tokens": 46787402.0,
      "step": 26975
    },
    {
      "entropy": 5.723541975021362,
      "epoch": 2.0991246839136353,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045546876838513134,
      "loss": 5.0229,
      "mean_token_accuracy": 0.19740087836980819,
      "num_tokens": 46796254.0,
      "step": 26980
    },
    {
      "entropy": 5.657139873504638,
      "epoch": 2.099513713285353,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045545221493396805,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20464279502630234,
      "num_tokens": 46804979.0,
      "step": 26985
    },
    {
      "entropy": 5.610894441604614,
      "epoch": 2.0999027426570707,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045543565874472963,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21238506883382796,
      "num_tokens": 46813718.0,
      "step": 26990
    },
    {
      "entropy": 5.701339960098267,
      "epoch": 2.1002917720287884,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045541909981766725,
      "loss": 5.0041,
      "mean_token_accuracy": 0.19869070798158645,
      "num_tokens": 46822265.0,
      "step": 26995
    },
    {
      "entropy": 5.658788871765137,
      "epoch": 2.1006808014005056,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004554025381530322,
      "loss": 4.9648,
      "mean_token_accuracy": 0.20687190741300582,
      "num_tokens": 46831292.0,
      "step": 27000
    },
    {
      "epoch": 2.1006808014005056,
      "eval_entropy": 5.398722856336195,
      "eval_loss": 5.089028358459473,
      "eval_mean_token_accuracy": 0.209127393660348,
      "eval_num_tokens": 46831292.0,
      "eval_runtime": 21.8497,
      "eval_samples_per_second": 1901.998,
      "eval_steps_per_second": 237.761,
      "step": 27000
    },
    {
      "entropy": 5.598618412017823,
      "epoch": 2.1010698307722233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004553859737510758,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20752410888671874,
      "num_tokens": 46840029.0,
      "step": 27005
    },
    {
      "entropy": 5.621729803085327,
      "epoch": 2.101458860143941,
      "grad_norm": 1.125,
      "learning_rate": 0.0004553694066120493,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21080188900232316,
      "num_tokens": 46848475.0,
      "step": 27010
    },
    {
      "entropy": 5.721709394454956,
      "epoch": 2.1018478895156583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045535283673620426,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20370381623506545,
      "num_tokens": 46857124.0,
      "step": 27015
    },
    {
      "entropy": 5.772908115386963,
      "epoch": 2.102236918887376,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045533626412379194,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1972390502691269,
      "num_tokens": 46865809.0,
      "step": 27020
    },
    {
      "entropy": 5.715380334854126,
      "epoch": 2.1026259482590937,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004553196887750638,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2008146181702614,
      "num_tokens": 46874845.0,
      "step": 27025
    },
    {
      "entropy": 5.648400831222534,
      "epoch": 2.103014977630811,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004553031106902714,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2061661198735237,
      "num_tokens": 46883740.0,
      "step": 27030
    },
    {
      "entropy": 5.618589782714844,
      "epoch": 2.1034040070025286,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045528652986966627,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2120565354824066,
      "num_tokens": 46891936.0,
      "step": 27035
    },
    {
      "entropy": 5.587720680236816,
      "epoch": 2.1037930363742463,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045526994631350006,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20316831320524215,
      "num_tokens": 46900918.0,
      "step": 27040
    },
    {
      "entropy": 5.740756177902222,
      "epoch": 2.104182065745964,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004552533600220243,
      "loss": 5.013,
      "mean_token_accuracy": 0.20343270599842073,
      "num_tokens": 46910785.0,
      "step": 27045
    },
    {
      "entropy": 5.762513256072998,
      "epoch": 2.1045710951176813,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045523677099549073,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20533704608678818,
      "num_tokens": 46919191.0,
      "step": 27050
    },
    {
      "entropy": 5.635636043548584,
      "epoch": 2.104960124489399,
      "grad_norm": 1.078125,
      "learning_rate": 0.000455220179234151,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20006966292858125,
      "num_tokens": 46927574.0,
      "step": 27055
    },
    {
      "entropy": 5.654203987121582,
      "epoch": 2.1053491538611167,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004552035847382569,
      "loss": 4.9236,
      "mean_token_accuracy": 0.20768364071846007,
      "num_tokens": 46935694.0,
      "step": 27060
    },
    {
      "entropy": 5.722375106811524,
      "epoch": 2.105738183232834,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045518698750806024,
      "loss": 5.0575,
      "mean_token_accuracy": 0.19995335191488267,
      "num_tokens": 46943929.0,
      "step": 27065
    },
    {
      "entropy": 5.640846824645996,
      "epoch": 2.1061272126045516,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045517038754381287,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2144556775689125,
      "num_tokens": 46952130.0,
      "step": 27070
    },
    {
      "entropy": 5.692103481292724,
      "epoch": 2.1065162419762693,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004551537848457666,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20545934289693832,
      "num_tokens": 46961010.0,
      "step": 27075
    },
    {
      "entropy": 5.6833172798156735,
      "epoch": 2.106905271347987,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004551371794141734,
      "loss": 4.994,
      "mean_token_accuracy": 0.20762139707803726,
      "num_tokens": 46969789.0,
      "step": 27080
    },
    {
      "entropy": 5.735100126266479,
      "epoch": 2.1072943007197042,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004551205712492852,
      "loss": 5.0257,
      "mean_token_accuracy": 0.20615569055080413,
      "num_tokens": 46977436.0,
      "step": 27085
    },
    {
      "entropy": 5.6994843006134035,
      "epoch": 2.107683330091422,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004551039603513541,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20444808304309844,
      "num_tokens": 46986065.0,
      "step": 27090
    },
    {
      "entropy": 5.654475259780884,
      "epoch": 2.1080723594631396,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045508734672063204,
      "loss": 5.0091,
      "mean_token_accuracy": 0.19613498598337173,
      "num_tokens": 46994355.0,
      "step": 27095
    },
    {
      "entropy": 5.628608322143554,
      "epoch": 2.108461388834857,
      "grad_norm": 1.125,
      "learning_rate": 0.0004550707303573711,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2078159585595131,
      "num_tokens": 47003028.0,
      "step": 27100
    },
    {
      "entropy": 5.7190276145935055,
      "epoch": 2.1088504182065746,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045505411126182347,
      "loss": 5.0854,
      "mean_token_accuracy": 0.19953883439302444,
      "num_tokens": 47012351.0,
      "step": 27105
    },
    {
      "entropy": 5.740923595428467,
      "epoch": 2.1092394475782923,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004550374894342412,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2128218799829483,
      "num_tokens": 47020600.0,
      "step": 27110
    },
    {
      "entropy": 5.6479698657989506,
      "epoch": 2.1096284769500095,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004550208648748767,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20195877999067308,
      "num_tokens": 47029276.0,
      "step": 27115
    },
    {
      "entropy": 5.647253322601318,
      "epoch": 2.1100175063217272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000455004237583982,
      "loss": 4.9333,
      "mean_token_accuracy": 0.20935706049203873,
      "num_tokens": 47037894.0,
      "step": 27120
    },
    {
      "entropy": 5.674368524551392,
      "epoch": 2.110406535693445,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004549876075618096,
      "loss": 4.9176,
      "mean_token_accuracy": 0.20982100665569306,
      "num_tokens": 47046602.0,
      "step": 27125
    },
    {
      "entropy": 5.713425731658935,
      "epoch": 2.110795565065162,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004549709748086117,
      "loss": 5.0656,
      "mean_token_accuracy": 0.19568216800689697,
      "num_tokens": 47056091.0,
      "step": 27130
    },
    {
      "entropy": 5.699641418457031,
      "epoch": 2.11118459443688,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045495433932464063,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19898207485675812,
      "num_tokens": 47065420.0,
      "step": 27135
    },
    {
      "entropy": 5.709368181228638,
      "epoch": 2.1115736238085976,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000454937701110149,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20001771301031113,
      "num_tokens": 47074247.0,
      "step": 27140
    },
    {
      "entropy": 5.778514814376831,
      "epoch": 2.1119626531803153,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045492106016538917,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19471174031496047,
      "num_tokens": 47083589.0,
      "step": 27145
    },
    {
      "entropy": 5.684567832946778,
      "epoch": 2.1123516825520325,
      "grad_norm": 1.25,
      "learning_rate": 0.00045490441649061354,
      "loss": 4.9527,
      "mean_token_accuracy": 0.20498155653476716,
      "num_tokens": 47091861.0,
      "step": 27150
    },
    {
      "entropy": 5.76125922203064,
      "epoch": 2.1127407119237502,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004548877700860747,
      "loss": 5.1442,
      "mean_token_accuracy": 0.19471305906772612,
      "num_tokens": 47101436.0,
      "step": 27155
    },
    {
      "entropy": 5.704738426208496,
      "epoch": 2.113129741295468,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045487112095202544,
      "loss": 4.9984,
      "mean_token_accuracy": 0.20823220014572144,
      "num_tokens": 47110390.0,
      "step": 27160
    },
    {
      "entropy": 5.748239898681641,
      "epoch": 2.113518770667185,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004548544690887181,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1970311611890793,
      "num_tokens": 47119001.0,
      "step": 27165
    },
    {
      "entropy": 5.629909944534302,
      "epoch": 2.113907800038903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004548378144964054,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20259853601455688,
      "num_tokens": 47127067.0,
      "step": 27170
    },
    {
      "entropy": 5.736306524276733,
      "epoch": 2.1142968294106206,
      "grad_norm": 1.125,
      "learning_rate": 0.00045482115717534024,
      "loss": 5.1455,
      "mean_token_accuracy": 0.19881599992513657,
      "num_tokens": 47136747.0,
      "step": 27175
    },
    {
      "entropy": 5.684993886947632,
      "epoch": 2.1146858587823383,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004548044971257752,
      "loss": 4.9488,
      "mean_token_accuracy": 0.2091785192489624,
      "num_tokens": 47145783.0,
      "step": 27180
    },
    {
      "entropy": 5.7073980331420895,
      "epoch": 2.1150748881540555,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045478783434796307,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2048191398382187,
      "num_tokens": 47154344.0,
      "step": 27185
    },
    {
      "entropy": 5.646864986419677,
      "epoch": 2.1154639175257732,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045477116884215675,
      "loss": 5.0105,
      "mean_token_accuracy": 0.20400911271572114,
      "num_tokens": 47163733.0,
      "step": 27190
    },
    {
      "entropy": 5.647171974182129,
      "epoch": 2.115852946897491,
      "grad_norm": 1.140625,
      "learning_rate": 0.000454754500608609,
      "loss": 4.9985,
      "mean_token_accuracy": 0.2054302766919136,
      "num_tokens": 47172341.0,
      "step": 27195
    },
    {
      "entropy": 5.553089952468872,
      "epoch": 2.116241976269208,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004547378296475729,
      "loss": 4.8066,
      "mean_token_accuracy": 0.22627432495355607,
      "num_tokens": 47181118.0,
      "step": 27200
    },
    {
      "entropy": 5.6924810886383055,
      "epoch": 2.116631005640926,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045472115595930124,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20456416457891463,
      "num_tokens": 47190348.0,
      "step": 27205
    },
    {
      "entropy": 5.686958265304566,
      "epoch": 2.1170200350126436,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045470447954404716,
      "loss": 5.0899,
      "mean_token_accuracy": 0.1961064726114273,
      "num_tokens": 47199812.0,
      "step": 27210
    },
    {
      "entropy": 5.691433763504028,
      "epoch": 2.117409064384361,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045468780040206364,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20367804318666458,
      "num_tokens": 47208547.0,
      "step": 27215
    },
    {
      "entropy": 5.691991901397705,
      "epoch": 2.1177980937560785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004546711185336037,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2120356321334839,
      "num_tokens": 47216980.0,
      "step": 27220
    },
    {
      "entropy": 5.703304290771484,
      "epoch": 2.1181871231277962,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045465443393892047,
      "loss": 5.0289,
      "mean_token_accuracy": 0.19839421808719634,
      "num_tokens": 47225701.0,
      "step": 27225
    },
    {
      "entropy": 5.682852649688721,
      "epoch": 2.118576152499514,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004546377466182673,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20499838292598724,
      "num_tokens": 47234760.0,
      "step": 27230
    },
    {
      "entropy": 5.767586040496826,
      "epoch": 2.118965181871231,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004546210565718972,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20037098973989487,
      "num_tokens": 47243392.0,
      "step": 27235
    },
    {
      "entropy": 5.717236709594727,
      "epoch": 2.119354211242949,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045460436380006337,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20907835513353348,
      "num_tokens": 47251297.0,
      "step": 27240
    },
    {
      "entropy": 5.604996967315674,
      "epoch": 2.1197432406146666,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004545876683030192,
      "loss": 4.899,
      "mean_token_accuracy": 0.20628928691148757,
      "num_tokens": 47259859.0,
      "step": 27245
    },
    {
      "entropy": 5.7250336647033695,
      "epoch": 2.120132269986384,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004545709700810181,
      "loss": 5.0411,
      "mean_token_accuracy": 0.2024758279323578,
      "num_tokens": 47268653.0,
      "step": 27250
    },
    {
      "entropy": 5.702297830581665,
      "epoch": 2.1205212993581015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004545542691343133,
      "loss": 4.9604,
      "mean_token_accuracy": 0.20446040183305741,
      "num_tokens": 47277254.0,
      "step": 27255
    },
    {
      "entropy": 5.693854331970215,
      "epoch": 2.120910328729819,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004545375654631582,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19718928784132003,
      "num_tokens": 47285885.0,
      "step": 27260
    },
    {
      "entropy": 5.660462474822998,
      "epoch": 2.1212993581015365,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045452085906780645,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20024797022342683,
      "num_tokens": 47294025.0,
      "step": 27265
    },
    {
      "entropy": 5.640049982070923,
      "epoch": 2.121688387473254,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045450414994851123,
      "loss": 4.9204,
      "mean_token_accuracy": 0.20874083489179612,
      "num_tokens": 47303531.0,
      "step": 27270
    },
    {
      "entropy": 5.715854501724243,
      "epoch": 2.122077416844972,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045448743810552636,
      "loss": 5.0718,
      "mean_token_accuracy": 0.19851782768964768,
      "num_tokens": 47311753.0,
      "step": 27275
    },
    {
      "entropy": 5.611560297012329,
      "epoch": 2.1224664462166896,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045447072353910537,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2008112907409668,
      "num_tokens": 47320736.0,
      "step": 27280
    },
    {
      "entropy": 5.734351873397827,
      "epoch": 2.122855475588407,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004544540062495016,
      "loss": 4.9977,
      "mean_token_accuracy": 0.2022951915860176,
      "num_tokens": 47329498.0,
      "step": 27285
    },
    {
      "entropy": 5.716881990432739,
      "epoch": 2.1232445049601245,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045443728623696905,
      "loss": 5.0226,
      "mean_token_accuracy": 0.20503806620836257,
      "num_tokens": 47338375.0,
      "step": 27290
    },
    {
      "entropy": 5.63335919380188,
      "epoch": 2.123633534331842,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045442056350176123,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20354518294334412,
      "num_tokens": 47346608.0,
      "step": 27295
    },
    {
      "entropy": 5.702572870254516,
      "epoch": 2.1240225637035595,
      "grad_norm": 1.296875,
      "learning_rate": 0.000454403838044132,
      "loss": 5.036,
      "mean_token_accuracy": 0.20045752227306365,
      "num_tokens": 47355375.0,
      "step": 27300
    },
    {
      "entropy": 5.705153703689575,
      "epoch": 2.124411593075277,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000454387109864335,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20235435962677,
      "num_tokens": 47363138.0,
      "step": 27305
    },
    {
      "entropy": 5.725976037979126,
      "epoch": 2.124800622446995,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045437037896262425,
      "loss": 5.0721,
      "mean_token_accuracy": 0.19554304629564284,
      "num_tokens": 47371431.0,
      "step": 27310
    },
    {
      "entropy": 5.632414627075195,
      "epoch": 2.125189651818712,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004543536453392534,
      "loss": 4.9738,
      "mean_token_accuracy": 0.21005242615938186,
      "num_tokens": 47379739.0,
      "step": 27315
    },
    {
      "entropy": 5.712195777893067,
      "epoch": 2.12557868119043,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045433690899447643,
      "loss": 5.0151,
      "mean_token_accuracy": 0.1999848797917366,
      "num_tokens": 47389240.0,
      "step": 27320
    },
    {
      "entropy": 5.714638137817383,
      "epoch": 2.1259677105621475,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045432016992854734,
      "loss": 4.9044,
      "mean_token_accuracy": 0.20921262055635453,
      "num_tokens": 47397666.0,
      "step": 27325
    },
    {
      "entropy": 5.689601135253906,
      "epoch": 2.126356739933865,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004543034281417201,
      "loss": 5.0362,
      "mean_token_accuracy": 0.19786131829023362,
      "num_tokens": 47407419.0,
      "step": 27330
    },
    {
      "entropy": 5.655070209503174,
      "epoch": 2.1267457693055825,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004542866836342488,
      "loss": 5.027,
      "mean_token_accuracy": 0.2031332567334175,
      "num_tokens": 47416750.0,
      "step": 27335
    },
    {
      "entropy": 5.697241926193238,
      "epoch": 2.1271347986773,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004542699364063873,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20288574546575547,
      "num_tokens": 47425833.0,
      "step": 27340
    },
    {
      "entropy": 5.669374465942383,
      "epoch": 2.127523828049018,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004542531864583899,
      "loss": 4.913,
      "mean_token_accuracy": 0.2076556697487831,
      "num_tokens": 47434470.0,
      "step": 27345
    },
    {
      "entropy": 5.700985717773437,
      "epoch": 2.127912857420735,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004542364337905108,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20635852664709092,
      "num_tokens": 47442904.0,
      "step": 27350
    },
    {
      "entropy": 5.685580158233643,
      "epoch": 2.128301886792453,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045421967840300403,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20642447471618652,
      "num_tokens": 47452373.0,
      "step": 27355
    },
    {
      "entropy": 5.669459819793701,
      "epoch": 2.1286909161641705,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004542029202961239,
      "loss": 4.8833,
      "mean_token_accuracy": 0.20824362933635712,
      "num_tokens": 47460504.0,
      "step": 27360
    },
    {
      "entropy": 5.6466676712036135,
      "epoch": 2.1290799455358878,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045418615947012464,
      "loss": 4.9809,
      "mean_token_accuracy": 0.20636244267225265,
      "num_tokens": 47469526.0,
      "step": 27365
    },
    {
      "entropy": 5.601554870605469,
      "epoch": 2.1294689749076054,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004541693959252607,
      "loss": 4.9311,
      "mean_token_accuracy": 0.20836524218320845,
      "num_tokens": 47478241.0,
      "step": 27370
    },
    {
      "entropy": 5.665599822998047,
      "epoch": 2.129858004279323,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045415262966178635,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21625545471906663,
      "num_tokens": 47486956.0,
      "step": 27375
    },
    {
      "entropy": 5.659033870697021,
      "epoch": 2.130247033651041,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004541358606799559,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22416906356811522,
      "num_tokens": 47494817.0,
      "step": 27380
    },
    {
      "entropy": 5.636460494995117,
      "epoch": 2.130636063022758,
      "grad_norm": 1.203125,
      "learning_rate": 0.000454119088980024,
      "loss": 4.9958,
      "mean_token_accuracy": 0.19951868504285813,
      "num_tokens": 47504124.0,
      "step": 27385
    },
    {
      "entropy": 5.722503280639648,
      "epoch": 2.131025092394476,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004541023145622449,
      "loss": 4.9685,
      "mean_token_accuracy": 0.20334541350603103,
      "num_tokens": 47513160.0,
      "step": 27390
    },
    {
      "entropy": 5.68713960647583,
      "epoch": 2.1314141217661935,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045408553742687336,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2097068503499031,
      "num_tokens": 47521298.0,
      "step": 27395
    },
    {
      "entropy": 5.628638982772827,
      "epoch": 2.1318031511379107,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004540687575741638,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21448117792606353,
      "num_tokens": 47529516.0,
      "step": 27400
    },
    {
      "entropy": 5.614070320129395,
      "epoch": 2.1321921805096284,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045405197500437077,
      "loss": 4.957,
      "mean_token_accuracy": 0.20356537252664567,
      "num_tokens": 47538311.0,
      "step": 27405
    },
    {
      "entropy": 5.765914535522461,
      "epoch": 2.132581209881346,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045403518971774915,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19366784691810607,
      "num_tokens": 47547186.0,
      "step": 27410
    },
    {
      "entropy": 5.664762115478515,
      "epoch": 2.132970239253064,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045401840171455343,
      "loss": 4.898,
      "mean_token_accuracy": 0.21105122715234756,
      "num_tokens": 47556173.0,
      "step": 27415
    },
    {
      "entropy": 5.662423038482666,
      "epoch": 2.133359268624781,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004540016109950384,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21242961883544922,
      "num_tokens": 47564595.0,
      "step": 27420
    },
    {
      "entropy": 5.604873085021973,
      "epoch": 2.133748297996499,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004539848175594589,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20666794627904891,
      "num_tokens": 47573559.0,
      "step": 27425
    },
    {
      "entropy": 5.658281946182251,
      "epoch": 2.1341373273682165,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004539680214080695,
      "loss": 4.8861,
      "mean_token_accuracy": 0.20779887437820435,
      "num_tokens": 47582178.0,
      "step": 27430
    },
    {
      "entropy": 5.624703216552734,
      "epoch": 2.1345263567399337,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045395122254112536,
      "loss": 4.8224,
      "mean_token_accuracy": 0.22041141241788864,
      "num_tokens": 47590738.0,
      "step": 27435
    },
    {
      "entropy": 5.776275157928467,
      "epoch": 2.1349153861116514,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004539344209588812,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19361840337514877,
      "num_tokens": 47599339.0,
      "step": 27440
    },
    {
      "entropy": 5.667601919174194,
      "epoch": 2.135304415483369,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045391761666159204,
      "loss": 4.946,
      "mean_token_accuracy": 0.2045337289571762,
      "num_tokens": 47608250.0,
      "step": 27445
    },
    {
      "entropy": 5.729393243789673,
      "epoch": 2.1356934448550864,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045390080964951294,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2063937157392502,
      "num_tokens": 47616823.0,
      "step": 27450
    },
    {
      "entropy": 5.695668935775757,
      "epoch": 2.136082474226804,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045388399992289864,
      "loss": 5.009,
      "mean_token_accuracy": 0.2079804241657257,
      "num_tokens": 47627002.0,
      "step": 27455
    },
    {
      "entropy": 5.836758470535278,
      "epoch": 2.136471503598522,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045386718748200436,
      "loss": 5.0797,
      "mean_token_accuracy": 0.2018369510769844,
      "num_tokens": 47635588.0,
      "step": 27460
    },
    {
      "entropy": 5.7780311584472654,
      "epoch": 2.136860532970239,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045385037232708534,
      "loss": 5.044,
      "mean_token_accuracy": 0.20847051292657853,
      "num_tokens": 47645089.0,
      "step": 27465
    },
    {
      "entropy": 5.7362385272979735,
      "epoch": 2.1372495623419567,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004538335544583964,
      "loss": 5.0366,
      "mean_token_accuracy": 0.1971965864300728,
      "num_tokens": 47654088.0,
      "step": 27470
    },
    {
      "entropy": 5.710936546325684,
      "epoch": 2.1376385917136744,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045381673387619306,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19516152888536453,
      "num_tokens": 47662367.0,
      "step": 27475
    },
    {
      "entropy": 5.669755315780639,
      "epoch": 2.138027621085392,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045379991058073035,
      "loss": 4.9901,
      "mean_token_accuracy": 0.19725206941366197,
      "num_tokens": 47670197.0,
      "step": 27480
    },
    {
      "entropy": 5.691903066635132,
      "epoch": 2.1384166504571094,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004537830845722636,
      "loss": 5.0559,
      "mean_token_accuracy": 0.2022085428237915,
      "num_tokens": 47678396.0,
      "step": 27485
    },
    {
      "entropy": 5.635660600662232,
      "epoch": 2.138805679828827,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004537662558510481,
      "loss": 4.9173,
      "mean_token_accuracy": 0.20544525980949402,
      "num_tokens": 47686671.0,
      "step": 27490
    },
    {
      "entropy": 5.710362577438355,
      "epoch": 2.1391947092005448,
      "grad_norm": 1.25,
      "learning_rate": 0.00045374942441733925,
      "loss": 5.0113,
      "mean_token_accuracy": 0.19232126623392104,
      "num_tokens": 47694923.0,
      "step": 27495
    },
    {
      "entropy": 5.661003971099854,
      "epoch": 2.139583738572262,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004537325902713923,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2042442336678505,
      "num_tokens": 47702944.0,
      "step": 27500
    },
    {
      "entropy": 5.701382684707641,
      "epoch": 2.1399727679439797,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004537157534134629,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21267200112342835,
      "num_tokens": 47711075.0,
      "step": 27505
    },
    {
      "entropy": 5.656701326370239,
      "epoch": 2.1403617973156974,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045369891384380623,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2053896814584732,
      "num_tokens": 47719338.0,
      "step": 27510
    },
    {
      "entropy": 5.697814846038819,
      "epoch": 2.140750826687415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004536820715626781,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20261289328336715,
      "num_tokens": 47728035.0,
      "step": 27515
    },
    {
      "entropy": 5.688754606246948,
      "epoch": 2.1411398560591324,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045366522657033385,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20124449133872985,
      "num_tokens": 47736438.0,
      "step": 27520
    },
    {
      "entropy": 5.679811716079712,
      "epoch": 2.14152888543085,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045364837886702924,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20542293041944504,
      "num_tokens": 47744661.0,
      "step": 27525
    },
    {
      "entropy": 5.655055141448974,
      "epoch": 2.1419179148025678,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004536315284530198,
      "loss": 4.8999,
      "mean_token_accuracy": 0.20525434166193007,
      "num_tokens": 47752550.0,
      "step": 27530
    },
    {
      "entropy": 5.673747873306274,
      "epoch": 2.142306944174285,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004536146753285612,
      "loss": 4.9698,
      "mean_token_accuracy": 0.2125828042626381,
      "num_tokens": 47761116.0,
      "step": 27535
    },
    {
      "entropy": 5.72344822883606,
      "epoch": 2.1426959735460027,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004535978194939093,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20122429579496384,
      "num_tokens": 47769804.0,
      "step": 27540
    },
    {
      "entropy": 5.658901357650757,
      "epoch": 2.1430850029177204,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045358096094931966,
      "loss": 4.938,
      "mean_token_accuracy": 0.20771819204092026,
      "num_tokens": 47777684.0,
      "step": 27545
    },
    {
      "entropy": 5.650033044815063,
      "epoch": 2.1434740322894377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004535640996950482,
      "loss": 4.9166,
      "mean_token_accuracy": 0.20991616696119308,
      "num_tokens": 47786443.0,
      "step": 27550
    },
    {
      "entropy": 5.682796430587769,
      "epoch": 2.1438630616611554,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004535472357313507,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20119493305683137,
      "num_tokens": 47795408.0,
      "step": 27555
    },
    {
      "entropy": 5.6264105319976805,
      "epoch": 2.144252091032873,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045353036905848316,
      "loss": 4.9166,
      "mean_token_accuracy": 0.20748237371444703,
      "num_tokens": 47804108.0,
      "step": 27560
    },
    {
      "entropy": 5.677252101898193,
      "epoch": 2.1446411204045903,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004535134996767014,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20655613839626313,
      "num_tokens": 47811972.0,
      "step": 27565
    },
    {
      "entropy": 5.688304376602173,
      "epoch": 2.145030149776308,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004534966275862614,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19587276726961136,
      "num_tokens": 47820615.0,
      "step": 27570
    },
    {
      "entropy": 5.720305490493774,
      "epoch": 2.1454191791480257,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004534797527874192,
      "loss": 5.0594,
      "mean_token_accuracy": 0.19924334138631822,
      "num_tokens": 47828773.0,
      "step": 27575
    },
    {
      "entropy": 5.675986671447754,
      "epoch": 2.1458082085197434,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045346287528043083,
      "loss": 4.9602,
      "mean_token_accuracy": 0.20447747707366942,
      "num_tokens": 47837197.0,
      "step": 27580
    },
    {
      "entropy": 5.679257106781006,
      "epoch": 2.1461972378914607,
      "grad_norm": 1.125,
      "learning_rate": 0.00045344599506555244,
      "loss": 5.1117,
      "mean_token_accuracy": 0.19356755167245865,
      "num_tokens": 47846127.0,
      "step": 27585
    },
    {
      "entropy": 5.696824550628662,
      "epoch": 2.1465862672631784,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045342911214304,
      "loss": 4.9772,
      "mean_token_accuracy": 0.21097922176122666,
      "num_tokens": 47854832.0,
      "step": 27590
    },
    {
      "entropy": 5.7173456192016605,
      "epoch": 2.146975296634896,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004534122265131498,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2136561632156372,
      "num_tokens": 47862621.0,
      "step": 27595
    },
    {
      "entropy": 5.706149530410767,
      "epoch": 2.1473643260066133,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045339533817613807,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19863034337759017,
      "num_tokens": 47871406.0,
      "step": 27600
    },
    {
      "entropy": 5.62992672920227,
      "epoch": 2.147753355378331,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045337844713226106,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21036789566278458,
      "num_tokens": 47880361.0,
      "step": 27605
    },
    {
      "entropy": 5.6924741744995115,
      "epoch": 2.1481423847500487,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045336155338177497,
      "loss": 5.0401,
      "mean_token_accuracy": 0.19467492401599884,
      "num_tokens": 47889632.0,
      "step": 27610
    },
    {
      "entropy": 5.682030391693115,
      "epoch": 2.1485314141217664,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004533446569249362,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21008805483579635,
      "num_tokens": 47898612.0,
      "step": 27615
    },
    {
      "entropy": 5.650802087783814,
      "epoch": 2.1489204434934837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045332775776200116,
      "loss": 5.0017,
      "mean_token_accuracy": 0.20346229374408722,
      "num_tokens": 47906628.0,
      "step": 27620
    },
    {
      "entropy": 5.611424541473388,
      "epoch": 2.1493094728652014,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045331085589322626,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21035341620445253,
      "num_tokens": 47915831.0,
      "step": 27625
    },
    {
      "entropy": 5.7616614818573,
      "epoch": 2.149698502236919,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004532939513188679,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20639760345220565,
      "num_tokens": 47924289.0,
      "step": 27630
    },
    {
      "entropy": 5.747340536117553,
      "epoch": 2.1500875316086363,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004532770440391826,
      "loss": 5.0934,
      "mean_token_accuracy": 0.19850980937480928,
      "num_tokens": 47932438.0,
      "step": 27635
    },
    {
      "entropy": 5.713750982284546,
      "epoch": 2.150476560980354,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004532601340544269,
      "loss": 4.9722,
      "mean_token_accuracy": 0.2006446212530136,
      "num_tokens": 47940634.0,
      "step": 27640
    },
    {
      "entropy": 5.692246627807617,
      "epoch": 2.1508655903520717,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004532432213648574,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20543272346258162,
      "num_tokens": 47949871.0,
      "step": 27645
    },
    {
      "entropy": 5.681021118164063,
      "epoch": 2.151254619723789,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045322630597073067,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21011193245649337,
      "num_tokens": 47957847.0,
      "step": 27650
    },
    {
      "entropy": 5.66308856010437,
      "epoch": 2.1516436490955066,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045320938787230355,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2075476363301277,
      "num_tokens": 47965978.0,
      "step": 27655
    },
    {
      "entropy": 5.75488772392273,
      "epoch": 2.1520326784672243,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045319246706983257,
      "loss": 5.1269,
      "mean_token_accuracy": 0.1963630810379982,
      "num_tokens": 47975407.0,
      "step": 27660
    },
    {
      "entropy": 5.685040378570557,
      "epoch": 2.152421707838942,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045317554356357446,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2093166470527649,
      "num_tokens": 47983652.0,
      "step": 27665
    },
    {
      "entropy": 5.730643796920776,
      "epoch": 2.1528107372106593,
      "grad_norm": 1.125,
      "learning_rate": 0.00045315861735378615,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20665504783391953,
      "num_tokens": 47992750.0,
      "step": 27670
    },
    {
      "entropy": 5.62513108253479,
      "epoch": 2.153199766582377,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045314168844072435,
      "loss": 4.9238,
      "mean_token_accuracy": 0.20908937454223633,
      "num_tokens": 48000930.0,
      "step": 27675
    },
    {
      "entropy": 5.634348821640015,
      "epoch": 2.1535887959540947,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045312475682464604,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20412052869796754,
      "num_tokens": 48009285.0,
      "step": 27680
    },
    {
      "entropy": 5.7203456401824955,
      "epoch": 2.153977825325812,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045310782250580794,
      "loss": 5.0542,
      "mean_token_accuracy": 0.19570087045431137,
      "num_tokens": 48017806.0,
      "step": 27685
    },
    {
      "entropy": 5.644450378417969,
      "epoch": 2.1543668546975296,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004530908854844672,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1966041460633278,
      "num_tokens": 48026930.0,
      "step": 27690
    },
    {
      "entropy": 5.714881038665771,
      "epoch": 2.1547558840692473,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045307394576088076,
      "loss": 5.0996,
      "mean_token_accuracy": 0.19602345079183578,
      "num_tokens": 48036039.0,
      "step": 27695
    },
    {
      "entropy": 5.682121610641479,
      "epoch": 2.1551449134409646,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004530570033353056,
      "loss": 4.9064,
      "mean_token_accuracy": 0.21640757620334625,
      "num_tokens": 48045278.0,
      "step": 27700
    },
    {
      "entropy": 5.703929471969604,
      "epoch": 2.1555339428126823,
      "grad_norm": 1.25,
      "learning_rate": 0.00045304005820799874,
      "loss": 5.0175,
      "mean_token_accuracy": 0.19960139840841293,
      "num_tokens": 48054448.0,
      "step": 27705
    },
    {
      "entropy": 5.7509623050689695,
      "epoch": 2.1559229721844,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004530231103792175,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19853113889694213,
      "num_tokens": 48062672.0,
      "step": 27710
    },
    {
      "entropy": 5.714931488037109,
      "epoch": 2.1563120015561177,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004530061598492188,
      "loss": 4.9474,
      "mean_token_accuracy": 0.2034710794687271,
      "num_tokens": 48072140.0,
      "step": 27715
    },
    {
      "entropy": 5.7365138053894045,
      "epoch": 2.156701030927835,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045298920661826004,
      "loss": 5.1162,
      "mean_token_accuracy": 0.18935914039611818,
      "num_tokens": 48080394.0,
      "step": 27720
    },
    {
      "entropy": 5.709145355224609,
      "epoch": 2.1570900602995526,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004529722506865984,
      "loss": 4.9985,
      "mean_token_accuracy": 0.2054038092494011,
      "num_tokens": 48088947.0,
      "step": 27725
    },
    {
      "entropy": 5.6714941501617435,
      "epoch": 2.1574790896712703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000452955292054491,
      "loss": 4.9557,
      "mean_token_accuracy": 0.20117945969104767,
      "num_tokens": 48097156.0,
      "step": 27730
    },
    {
      "entropy": 5.691969728469848,
      "epoch": 2.1578681190429876,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045293833072219535,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2128015175461769,
      "num_tokens": 48105692.0,
      "step": 27735
    },
    {
      "entropy": 5.683024978637695,
      "epoch": 2.1582571484147053,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045292136668996876,
      "loss": 4.9737,
      "mean_token_accuracy": 0.20566923469305037,
      "num_tokens": 48114967.0,
      "step": 27740
    },
    {
      "entropy": 5.662674808502198,
      "epoch": 2.158646177786423,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004529043999580686,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2033538892865181,
      "num_tokens": 48123933.0,
      "step": 27745
    },
    {
      "entropy": 5.68277063369751,
      "epoch": 2.1590352071581402,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045288743052675234,
      "loss": 4.9619,
      "mean_token_accuracy": 0.2093115597963333,
      "num_tokens": 48131998.0,
      "step": 27750
    },
    {
      "entropy": 5.666032075881958,
      "epoch": 2.159424236529858,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045287045839627744,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2111566886305809,
      "num_tokens": 48140111.0,
      "step": 27755
    },
    {
      "entropy": 5.690597295761108,
      "epoch": 2.1598132659015756,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045285348356690155,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20501011312007905,
      "num_tokens": 48148826.0,
      "step": 27760
    },
    {
      "entropy": 5.697222566604614,
      "epoch": 2.1602022952732933,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000452836506038882,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20442713797092438,
      "num_tokens": 48157242.0,
      "step": 27765
    },
    {
      "entropy": 5.739878940582275,
      "epoch": 2.1605913246450106,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045281952581247654,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1970132514834404,
      "num_tokens": 48166571.0,
      "step": 27770
    },
    {
      "entropy": 5.717183303833008,
      "epoch": 2.1609803540167283,
      "grad_norm": 1.25,
      "learning_rate": 0.00045280254288794286,
      "loss": 4.9871,
      "mean_token_accuracy": 0.21210711747407912,
      "num_tokens": 48175280.0,
      "step": 27775
    },
    {
      "entropy": 5.723680925369263,
      "epoch": 2.161369383388446,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045278555726553855,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20551867634058,
      "num_tokens": 48184931.0,
      "step": 27780
    },
    {
      "entropy": 5.634449243545532,
      "epoch": 2.1617584127601632,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045276856894552143,
      "loss": 4.9197,
      "mean_token_accuracy": 0.20997531414031984,
      "num_tokens": 48192945.0,
      "step": 27785
    },
    {
      "entropy": 5.687900161743164,
      "epoch": 2.162147442131881,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004527515779281492,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20005444586277008,
      "num_tokens": 48201963.0,
      "step": 27790
    },
    {
      "entropy": 5.696456384658814,
      "epoch": 2.1625364715035986,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045273458421367976,
      "loss": 5.026,
      "mean_token_accuracy": 0.20505751967430114,
      "num_tokens": 48211178.0,
      "step": 27795
    },
    {
      "entropy": 5.68576831817627,
      "epoch": 2.162925500875316,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004527175878023708,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2098630040884018,
      "num_tokens": 48219294.0,
      "step": 27800
    },
    {
      "entropy": 5.616510343551636,
      "epoch": 2.1633145302470336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004527005886944803,
      "loss": 4.9805,
      "mean_token_accuracy": 0.20231364965438842,
      "num_tokens": 48227782.0,
      "step": 27805
    },
    {
      "entropy": 5.749646854400635,
      "epoch": 2.1637035596187513,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045268358689026626,
      "loss": 5.0028,
      "mean_token_accuracy": 0.19973407089710235,
      "num_tokens": 48236850.0,
      "step": 27810
    },
    {
      "entropy": 5.76823935508728,
      "epoch": 2.164092588990469,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004526665823899866,
      "loss": 5.0603,
      "mean_token_accuracy": 0.19898138344287872,
      "num_tokens": 48246095.0,
      "step": 27815
    },
    {
      "entropy": 5.7417655944824215,
      "epoch": 2.164481618362186,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045264957519389936,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2048045814037323,
      "num_tokens": 48254943.0,
      "step": 27820
    },
    {
      "entropy": 5.670479249954224,
      "epoch": 2.164870647733904,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045263256530226253,
      "loss": 4.9297,
      "mean_token_accuracy": 0.20750837922096252,
      "num_tokens": 48263954.0,
      "step": 27825
    },
    {
      "entropy": 5.674408388137818,
      "epoch": 2.1652596771056216,
      "grad_norm": 1.25,
      "learning_rate": 0.0004526155527153343,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20142923593521117,
      "num_tokens": 48272708.0,
      "step": 27830
    },
    {
      "entropy": 5.748548698425293,
      "epoch": 2.165648706477339,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004525985374333727,
      "loss": 5.0476,
      "mean_token_accuracy": 0.19351772367954254,
      "num_tokens": 48280782.0,
      "step": 27835
    },
    {
      "entropy": 5.663448762893677,
      "epoch": 2.1660377358490566,
      "grad_norm": 1.171875,
      "learning_rate": 0.000452581519456636,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2067514330148697,
      "num_tokens": 48289105.0,
      "step": 27840
    },
    {
      "entropy": 5.643997383117676,
      "epoch": 2.1664267652207743,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045256449878538243,
      "loss": 4.9847,
      "mean_token_accuracy": 0.2099043384194374,
      "num_tokens": 48297552.0,
      "step": 27845
    },
    {
      "entropy": 5.6354522705078125,
      "epoch": 2.166815794592492,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045254747541987017,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2046032190322876,
      "num_tokens": 48307541.0,
      "step": 27850
    },
    {
      "entropy": 5.703324794769287,
      "epoch": 2.167204823964209,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004525304493603576,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20304527431726455,
      "num_tokens": 48316078.0,
      "step": 27855
    },
    {
      "entropy": 5.742888259887695,
      "epoch": 2.167593853335927,
      "grad_norm": 1.25,
      "learning_rate": 0.00045251342060710295,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20033112913370132,
      "num_tokens": 48324221.0,
      "step": 27860
    },
    {
      "entropy": 5.732780504226684,
      "epoch": 2.1679828827076446,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004524963891603647,
      "loss": 5.1165,
      "mean_token_accuracy": 0.20213199257850648,
      "num_tokens": 48333047.0,
      "step": 27865
    },
    {
      "entropy": 5.7075590133667,
      "epoch": 2.168371912079362,
      "grad_norm": 1.125,
      "learning_rate": 0.00045247935502040136,
      "loss": 5.0765,
      "mean_token_accuracy": 0.20395572781562804,
      "num_tokens": 48342255.0,
      "step": 27870
    },
    {
      "entropy": 5.677798938751221,
      "epoch": 2.1687609414510796,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004524623181874712,
      "loss": 5.0794,
      "mean_token_accuracy": 0.19253385066986084,
      "num_tokens": 48350877.0,
      "step": 27875
    },
    {
      "entropy": 5.794387340545654,
      "epoch": 2.1691499708227973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004524452786618329,
      "loss": 5.1772,
      "mean_token_accuracy": 0.18598457872867585,
      "num_tokens": 48360537.0,
      "step": 27880
    },
    {
      "entropy": 5.743813133239746,
      "epoch": 2.1695390001945145,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045242823644374484,
      "loss": 5.0932,
      "mean_token_accuracy": 0.19653849750757219,
      "num_tokens": 48370518.0,
      "step": 27885
    },
    {
      "entropy": 5.674924373626709,
      "epoch": 2.169928029566232,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004524111915334658,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2108231395483017,
      "num_tokens": 48379379.0,
      "step": 27890
    },
    {
      "entropy": 5.6289306640625,
      "epoch": 2.17031705893795,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045239414393125424,
      "loss": 4.889,
      "mean_token_accuracy": 0.20944453477859498,
      "num_tokens": 48388090.0,
      "step": 27895
    },
    {
      "entropy": 5.742839002609253,
      "epoch": 2.170706088309667,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004523770936373689,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1943758472800255,
      "num_tokens": 48396527.0,
      "step": 27900
    },
    {
      "entropy": 5.710853528976441,
      "epoch": 2.171095117681385,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004523600406520685,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20709913820028306,
      "num_tokens": 48404632.0,
      "step": 27905
    },
    {
      "entropy": 5.620260286331177,
      "epoch": 2.1714841470531026,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004523429849756118,
      "loss": 4.9183,
      "mean_token_accuracy": 0.21835287809371948,
      "num_tokens": 48413573.0,
      "step": 27910
    },
    {
      "entropy": 5.750569152832031,
      "epoch": 2.1718731764248203,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004523259266082576,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1987303227186203,
      "num_tokens": 48421252.0,
      "step": 27915
    },
    {
      "entropy": 5.72512001991272,
      "epoch": 2.1722622057965375,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045230886555026455,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20141152888536454,
      "num_tokens": 48430322.0,
      "step": 27920
    },
    {
      "entropy": 5.729338788986206,
      "epoch": 2.172651235168255,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045229180180189176,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2137909084558487,
      "num_tokens": 48439570.0,
      "step": 27925
    },
    {
      "entropy": 5.6773584365844725,
      "epoch": 2.173040264539973,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045227473536339817,
      "loss": 5.0109,
      "mean_token_accuracy": 0.1942712187767029,
      "num_tokens": 48447803.0,
      "step": 27930
    },
    {
      "entropy": 5.7015077590942385,
      "epoch": 2.17342929391169,
      "grad_norm": 1.25,
      "learning_rate": 0.0004522576662350425,
      "loss": 5.0604,
      "mean_token_accuracy": 0.2004782423377037,
      "num_tokens": 48456694.0,
      "step": 27935
    },
    {
      "entropy": 5.708668613433838,
      "epoch": 2.173818323283408,
      "grad_norm": 1.3125,
      "learning_rate": 0.000452240594417084,
      "loss": 5.0543,
      "mean_token_accuracy": 0.20268608033657073,
      "num_tokens": 48465321.0,
      "step": 27940
    },
    {
      "entropy": 5.722630500793457,
      "epoch": 2.1742073526551255,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004522235199097815,
      "loss": 5.0421,
      "mean_token_accuracy": 0.20109671503305435,
      "num_tokens": 48474299.0,
      "step": 27945
    },
    {
      "entropy": 5.628232526779175,
      "epoch": 2.1745963820268432,
      "grad_norm": 1.125,
      "learning_rate": 0.0004522064427133942,
      "loss": 4.9292,
      "mean_token_accuracy": 0.20780753940343857,
      "num_tokens": 48483427.0,
      "step": 27950
    },
    {
      "entropy": 5.669827651977539,
      "epoch": 2.1749854113985605,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045218936282818116,
      "loss": 4.9174,
      "mean_token_accuracy": 0.20949266403913497,
      "num_tokens": 48492610.0,
      "step": 27955
    },
    {
      "entropy": 5.689406871795654,
      "epoch": 2.175374440770278,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004521722802544016,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2001194402575493,
      "num_tokens": 48502531.0,
      "step": 27960
    },
    {
      "entropy": 5.772637033462525,
      "epoch": 2.175763470141996,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045215519499231465,
      "loss": 5.007,
      "mean_token_accuracy": 0.2012021616101265,
      "num_tokens": 48511167.0,
      "step": 27965
    },
    {
      "entropy": 5.754801273345947,
      "epoch": 2.176152499513713,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045213810704217963,
      "loss": 5.0658,
      "mean_token_accuracy": 0.20177143961191177,
      "num_tokens": 48520355.0,
      "step": 27970
    },
    {
      "entropy": 5.665313673019409,
      "epoch": 2.176541528885431,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004521210164042557,
      "loss": 5.0464,
      "mean_token_accuracy": 0.2039900988340378,
      "num_tokens": 48529269.0,
      "step": 27975
    },
    {
      "entropy": 5.667738056182861,
      "epoch": 2.1769305582571485,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004521039230788024,
      "loss": 4.9405,
      "mean_token_accuracy": 0.20933383852243423,
      "num_tokens": 48537764.0,
      "step": 27980
    },
    {
      "entropy": 5.64856858253479,
      "epoch": 2.177319587628866,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004520868270660789,
      "loss": 4.995,
      "mean_token_accuracy": 0.2030662253499031,
      "num_tokens": 48547168.0,
      "step": 27985
    },
    {
      "entropy": 5.620655822753906,
      "epoch": 2.1777086170005835,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004520697283663446,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2040308192372322,
      "num_tokens": 48555809.0,
      "step": 27990
    },
    {
      "entropy": 5.749118852615356,
      "epoch": 2.178097646372301,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045205262697985897,
      "loss": 5.0237,
      "mean_token_accuracy": 0.19888675659894944,
      "num_tokens": 48564754.0,
      "step": 27995
    },
    {
      "entropy": 5.657030916213989,
      "epoch": 2.1784866757440184,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004520355229068816,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20034458637237548,
      "num_tokens": 48573732.0,
      "step": 28000
    },
    {
      "entropy": 5.658715772628784,
      "epoch": 2.178875705115736,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045201841614767203,
      "loss": 4.9189,
      "mean_token_accuracy": 0.20989740043878555,
      "num_tokens": 48581676.0,
      "step": 28005
    },
    {
      "entropy": 5.652215528488159,
      "epoch": 2.179264734487454,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045200130670248974,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20053718239068985,
      "num_tokens": 48590310.0,
      "step": 28010
    },
    {
      "entropy": 5.6476537704467775,
      "epoch": 2.1796537638591715,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045198419457159424,
      "loss": 4.9935,
      "mean_token_accuracy": 0.2088113009929657,
      "num_tokens": 48599306.0,
      "step": 28015
    },
    {
      "entropy": 5.802316951751709,
      "epoch": 2.180042793230889,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004519670797552453,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19432310611009598,
      "num_tokens": 48608264.0,
      "step": 28020
    },
    {
      "entropy": 5.726743984222412,
      "epoch": 2.1804318226026065,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045194996225370266,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20745579302310943,
      "num_tokens": 48617102.0,
      "step": 28025
    },
    {
      "entropy": 5.727710580825805,
      "epoch": 2.180820851974324,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045193284206722595,
      "loss": 5.0782,
      "mean_token_accuracy": 0.19640626907348632,
      "num_tokens": 48626207.0,
      "step": 28030
    },
    {
      "entropy": 5.6992645263671875,
      "epoch": 2.1812098813460414,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004519157191960749,
      "loss": 4.9179,
      "mean_token_accuracy": 0.20697971284389496,
      "num_tokens": 48634322.0,
      "step": 28035
    },
    {
      "entropy": 5.670881891250611,
      "epoch": 2.181598910717759,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004518985936405096,
      "loss": 4.9518,
      "mean_token_accuracy": 0.2139478802680969,
      "num_tokens": 48642026.0,
      "step": 28040
    },
    {
      "entropy": 5.6415602684021,
      "epoch": 2.181987940089477,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004518814654007896,
      "loss": 4.8856,
      "mean_token_accuracy": 0.21168052405118942,
      "num_tokens": 48650114.0,
      "step": 28045
    },
    {
      "entropy": 5.736222934722901,
      "epoch": 2.1823769694611945,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045186433447717486,
      "loss": 5.0208,
      "mean_token_accuracy": 0.19912574887275697,
      "num_tokens": 48660185.0,
      "step": 28050
    },
    {
      "entropy": 5.768896865844726,
      "epoch": 2.182765998832912,
      "grad_norm": 1.25,
      "learning_rate": 0.00045184720086992536,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20345480740070343,
      "num_tokens": 48668603.0,
      "step": 28055
    },
    {
      "entropy": 5.713300561904907,
      "epoch": 2.1831550282046295,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045183006457930107,
      "loss": 5.0327,
      "mean_token_accuracy": 0.20257086008787156,
      "num_tokens": 48676962.0,
      "step": 28060
    },
    {
      "entropy": 5.675505781173706,
      "epoch": 2.183544057576347,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004518129256055621,
      "loss": 5.022,
      "mean_token_accuracy": 0.20534039586782454,
      "num_tokens": 48685823.0,
      "step": 28065
    },
    {
      "entropy": 5.652837562561035,
      "epoch": 2.1839330869480644,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004517957839489683,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21189606040716172,
      "num_tokens": 48695116.0,
      "step": 28070
    },
    {
      "entropy": 5.667135429382324,
      "epoch": 2.184322116319782,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045177863960977986,
      "loss": 4.994,
      "mean_token_accuracy": 0.2040344163775444,
      "num_tokens": 48702725.0,
      "step": 28075
    },
    {
      "entropy": 5.651153564453125,
      "epoch": 2.1847111456915,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000451761492588257,
      "loss": 4.999,
      "mean_token_accuracy": 0.21182008385658263,
      "num_tokens": 48711104.0,
      "step": 28080
    },
    {
      "entropy": 5.661631155014038,
      "epoch": 2.185100175063217,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045174434288465986,
      "loss": 4.9652,
      "mean_token_accuracy": 0.20164453536272048,
      "num_tokens": 48720177.0,
      "step": 28085
    },
    {
      "entropy": 5.654524326324463,
      "epoch": 2.1854892044349348,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004517271904992486,
      "loss": 4.9335,
      "mean_token_accuracy": 0.20116988569498062,
      "num_tokens": 48728729.0,
      "step": 28090
    },
    {
      "entropy": 5.631008100509644,
      "epoch": 2.1858782338066525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004517100354322835,
      "loss": 4.9362,
      "mean_token_accuracy": 0.20874652564525603,
      "num_tokens": 48736876.0,
      "step": 28095
    },
    {
      "entropy": 5.627217245101929,
      "epoch": 2.18626726317837,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045169287768402497,
      "loss": 4.9119,
      "mean_token_accuracy": 0.20984110236167908,
      "num_tokens": 48745957.0,
      "step": 28100
    },
    {
      "entropy": 5.797620058059692,
      "epoch": 2.1866562925500874,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004516757172547331,
      "loss": 5.0809,
      "mean_token_accuracy": 0.2026408553123474,
      "num_tokens": 48754488.0,
      "step": 28105
    },
    {
      "entropy": 5.730008602142334,
      "epoch": 2.187045321921805,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004516585541446685,
      "loss": 4.9712,
      "mean_token_accuracy": 0.20777266025543212,
      "num_tokens": 48762797.0,
      "step": 28110
    },
    {
      "entropy": 5.605029582977295,
      "epoch": 2.187434351293523,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045164138835409153,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20266563445329666,
      "num_tokens": 48771639.0,
      "step": 28115
    },
    {
      "entropy": 5.636875820159912,
      "epoch": 2.18782338066524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045162421988326266,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20492016673088073,
      "num_tokens": 48779978.0,
      "step": 28120
    },
    {
      "entropy": 5.693480491638184,
      "epoch": 2.1882124100369578,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004516070487324424,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19926518946886063,
      "num_tokens": 48788202.0,
      "step": 28125
    },
    {
      "entropy": 5.722810316085815,
      "epoch": 2.1886014394086755,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045158987490189124,
      "loss": 5.006,
      "mean_token_accuracy": 0.2019208252429962,
      "num_tokens": 48796533.0,
      "step": 28130
    },
    {
      "entropy": 5.72516279220581,
      "epoch": 2.1889904687803927,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004515726983918697,
      "loss": 5.0444,
      "mean_token_accuracy": 0.19517805576324462,
      "num_tokens": 48805627.0,
      "step": 28135
    },
    {
      "entropy": 5.738078975677491,
      "epoch": 2.1893794981521104,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045155551920263864,
      "loss": 5.0363,
      "mean_token_accuracy": 0.19804058372974395,
      "num_tokens": 48814132.0,
      "step": 28140
    },
    {
      "entropy": 5.7138372421264645,
      "epoch": 2.189768527523828,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045153833733445857,
      "loss": 4.9775,
      "mean_token_accuracy": 0.19936518371105194,
      "num_tokens": 48823210.0,
      "step": 28145
    },
    {
      "entropy": 5.687272644042968,
      "epoch": 2.190157556895546,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004515211527875903,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2034782811999321,
      "num_tokens": 48831835.0,
      "step": 28150
    },
    {
      "entropy": 5.7396559715271,
      "epoch": 2.190546586267263,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004515039655622943,
      "loss": 5.0553,
      "mean_token_accuracy": 0.19870362132787706,
      "num_tokens": 48840917.0,
      "step": 28155
    },
    {
      "entropy": 5.740657663345337,
      "epoch": 2.1909356156389808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045148677565883167,
      "loss": 4.9911,
      "mean_token_accuracy": 0.1979047477245331,
      "num_tokens": 48849328.0,
      "step": 28160
    },
    {
      "entropy": 5.659861278533936,
      "epoch": 2.1913246450106985,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004514695830774631,
      "loss": 4.9934,
      "mean_token_accuracy": 0.2037700355052948,
      "num_tokens": 48858511.0,
      "step": 28165
    },
    {
      "entropy": 5.699523162841797,
      "epoch": 2.1917136743824157,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045145238781844953,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21030766367912293,
      "num_tokens": 48867152.0,
      "step": 28170
    },
    {
      "entropy": 5.690017032623291,
      "epoch": 2.1921027037541334,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045143518988205184,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20330598801374436,
      "num_tokens": 48875233.0,
      "step": 28175
    },
    {
      "entropy": 5.627080011367798,
      "epoch": 2.192491733125851,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045141798926853094,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20209655463695525,
      "num_tokens": 48884569.0,
      "step": 28180
    },
    {
      "entropy": 5.713271379470825,
      "epoch": 2.1928807624975684,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004514007859781478,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19722182005643846,
      "num_tokens": 48894402.0,
      "step": 28185
    },
    {
      "entropy": 5.809972476959229,
      "epoch": 2.193269791869286,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004513835800111635,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20810288786888123,
      "num_tokens": 48902105.0,
      "step": 28190
    },
    {
      "entropy": 5.738625478744507,
      "epoch": 2.1936588212410038,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004513663713678392,
      "loss": 5.0951,
      "mean_token_accuracy": 0.19630369693040847,
      "num_tokens": 48911650.0,
      "step": 28195
    },
    {
      "entropy": 5.676010227203369,
      "epoch": 2.1940478506127215,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045134916004843596,
      "loss": 4.962,
      "mean_token_accuracy": 0.20384355187416076,
      "num_tokens": 48920035.0,
      "step": 28200
    },
    {
      "entropy": 5.683791255950927,
      "epoch": 2.1944368799844387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004513319460532148,
      "loss": 4.9597,
      "mean_token_accuracy": 0.20637443512678147,
      "num_tokens": 48928460.0,
      "step": 28205
    },
    {
      "entropy": 5.780375623703003,
      "epoch": 2.1948259093561564,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045131472938243707,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2063479542732239,
      "num_tokens": 48937606.0,
      "step": 28210
    },
    {
      "entropy": 5.6919067859649655,
      "epoch": 2.195214938727874,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004512975100363639,
      "loss": 4.8958,
      "mean_token_accuracy": 0.20787456929683684,
      "num_tokens": 48945782.0,
      "step": 28215
    },
    {
      "entropy": 5.698641967773438,
      "epoch": 2.1956039680995914,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045128028801525675,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20155792385339738,
      "num_tokens": 48954243.0,
      "step": 28220
    },
    {
      "entropy": 5.709740924835205,
      "epoch": 2.195992997471309,
      "grad_norm": 1.25,
      "learning_rate": 0.00045126306331937674,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20751579701900483,
      "num_tokens": 48962578.0,
      "step": 28225
    },
    {
      "entropy": 5.678360223770142,
      "epoch": 2.1963820268430267,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004512458359489853,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2079502522945404,
      "num_tokens": 48971237.0,
      "step": 28230
    },
    {
      "entropy": 5.673741388320923,
      "epoch": 2.196771056214744,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004512286059043437,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21112075597047805,
      "num_tokens": 48979954.0,
      "step": 28235
    },
    {
      "entropy": 5.748825883865356,
      "epoch": 2.1971600855864617,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045121137318571364,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19700155407190323,
      "num_tokens": 48989174.0,
      "step": 28240
    },
    {
      "entropy": 5.765066862106323,
      "epoch": 2.1975491149581794,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045119413779335645,
      "loss": 5.0204,
      "mean_token_accuracy": 0.20112432390451432,
      "num_tokens": 48998341.0,
      "step": 28245
    },
    {
      "entropy": 5.682904148101807,
      "epoch": 2.197938144329897,
      "grad_norm": 1.25,
      "learning_rate": 0.0004511768997275336,
      "loss": 4.9459,
      "mean_token_accuracy": 0.20823090821504592,
      "num_tokens": 49007024.0,
      "step": 28250
    },
    {
      "entropy": 5.642064762115479,
      "epoch": 2.1983271737016143,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004511596589885068,
      "loss": 5.0257,
      "mean_token_accuracy": 0.20023765563964843,
      "num_tokens": 49015632.0,
      "step": 28255
    },
    {
      "entropy": 5.676866292953491,
      "epoch": 2.198716203073332,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004511424155765375,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20983691960573198,
      "num_tokens": 49024056.0,
      "step": 28260
    },
    {
      "entropy": 5.651511335372925,
      "epoch": 2.1991052324450497,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045112516949188735,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2053225815296173,
      "num_tokens": 49033095.0,
      "step": 28265
    },
    {
      "entropy": 5.660645771026611,
      "epoch": 2.199494261816767,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004511079207348182,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2069698691368103,
      "num_tokens": 49041632.0,
      "step": 28270
    },
    {
      "entropy": 5.734529113769531,
      "epoch": 2.1998832911884847,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004510906693055916,
      "loss": 5.0586,
      "mean_token_accuracy": 0.19970477521419525,
      "num_tokens": 49050336.0,
      "step": 28275
    },
    {
      "entropy": 5.664906644821167,
      "epoch": 2.2002723205602024,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004510734152044694,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20711356401443481,
      "num_tokens": 49058567.0,
      "step": 28280
    },
    {
      "entropy": 5.649997806549072,
      "epoch": 2.20066134993192,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045105615843171336,
      "loss": 4.8419,
      "mean_token_accuracy": 0.21842979192733764,
      "num_tokens": 49067003.0,
      "step": 28285
    },
    {
      "entropy": 5.718736886978149,
      "epoch": 2.2010503793036373,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045103889898758534,
      "loss": 5.0151,
      "mean_token_accuracy": 0.19904953390359878,
      "num_tokens": 49074965.0,
      "step": 28290
    },
    {
      "entropy": 5.679871273040772,
      "epoch": 2.201439408675355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004510216368723472,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2057369127869606,
      "num_tokens": 49085106.0,
      "step": 28295
    },
    {
      "entropy": 5.731793355941773,
      "epoch": 2.2018284380470727,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045100437208626086,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1967105582356453,
      "num_tokens": 49093125.0,
      "step": 28300
    },
    {
      "entropy": 5.678661108016968,
      "epoch": 2.20221746741879,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045098710462958844,
      "loss": 4.9185,
      "mean_token_accuracy": 0.20839287489652633,
      "num_tokens": 49101956.0,
      "step": 28305
    },
    {
      "entropy": 5.663334751129151,
      "epoch": 2.2026064967905077,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045096983450259166,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2060643419623375,
      "num_tokens": 49110942.0,
      "step": 28310
    },
    {
      "entropy": 5.739475107192993,
      "epoch": 2.2029955261622254,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045095256170553285,
      "loss": 5.0793,
      "mean_token_accuracy": 0.2004253789782524,
      "num_tokens": 49119181.0,
      "step": 28315
    },
    {
      "entropy": 5.707302474975586,
      "epoch": 2.2033845555339426,
      "grad_norm": 1.25,
      "learning_rate": 0.00045093528623867384,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20926332622766494,
      "num_tokens": 49127451.0,
      "step": 28320
    },
    {
      "entropy": 5.724442815780639,
      "epoch": 2.2037735849056603,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045091800810227706,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2024417206645012,
      "num_tokens": 49135550.0,
      "step": 28325
    },
    {
      "entropy": 5.594084787368774,
      "epoch": 2.204162614277378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004509007272966044,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2108796015381813,
      "num_tokens": 49144551.0,
      "step": 28330
    },
    {
      "entropy": 5.709326505661011,
      "epoch": 2.2045516436490953,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004508834438219182,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20962454676628112,
      "num_tokens": 49153079.0,
      "step": 28335
    },
    {
      "entropy": 5.709550523757935,
      "epoch": 2.204940673020813,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004508661576784807,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20048341155052185,
      "num_tokens": 49161253.0,
      "step": 28340
    },
    {
      "entropy": 5.629749345779419,
      "epoch": 2.2053297023925307,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004508488688665541,
      "loss": 4.9574,
      "mean_token_accuracy": 0.20701866298913957,
      "num_tokens": 49170192.0,
      "step": 28345
    },
    {
      "entropy": 5.712760543823242,
      "epoch": 2.2057187317642484,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004508315773864009,
      "loss": 5.017,
      "mean_token_accuracy": 0.20108174681663513,
      "num_tokens": 49179512.0,
      "step": 28350
    },
    {
      "entropy": 5.705562257766724,
      "epoch": 2.2061077611359656,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045081428323828337,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2111862048506737,
      "num_tokens": 49188431.0,
      "step": 28355
    },
    {
      "entropy": 5.670009326934815,
      "epoch": 2.2064967905076833,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004507969864224639,
      "loss": 4.9452,
      "mean_token_accuracy": 0.20929117500782013,
      "num_tokens": 49196852.0,
      "step": 28360
    },
    {
      "entropy": 5.664839601516723,
      "epoch": 2.206885819879401,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000450779686939205,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2051306113600731,
      "num_tokens": 49206007.0,
      "step": 28365
    },
    {
      "entropy": 5.6994832992553714,
      "epoch": 2.2072748492511183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045076238478876916,
      "loss": 4.952,
      "mean_token_accuracy": 0.20666207373142242,
      "num_tokens": 49214016.0,
      "step": 28370
    },
    {
      "entropy": 5.65054087638855,
      "epoch": 2.207663878622836,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004507450799714188,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21339460760354995,
      "num_tokens": 49222558.0,
      "step": 28375
    },
    {
      "entropy": 5.687335968017578,
      "epoch": 2.2080529079945537,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045072777248741664,
      "loss": 5.0035,
      "mean_token_accuracy": 0.19511208683252335,
      "num_tokens": 49230983.0,
      "step": 28380
    },
    {
      "entropy": 5.756176614761353,
      "epoch": 2.2084419373662714,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004507104623370252,
      "loss": 5.0545,
      "mean_token_accuracy": 0.19733452796936035,
      "num_tokens": 49239429.0,
      "step": 28385
    },
    {
      "entropy": 5.71230845451355,
      "epoch": 2.2088309667379886,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004506931495205072,
      "loss": 5.0012,
      "mean_token_accuracy": 0.19810230135917664,
      "num_tokens": 49247316.0,
      "step": 28390
    },
    {
      "entropy": 5.655335235595703,
      "epoch": 2.2092199961097063,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004506758340381253,
      "loss": 4.9462,
      "mean_token_accuracy": 0.20672931373119355,
      "num_tokens": 49256293.0,
      "step": 28395
    },
    {
      "entropy": 5.777633905410767,
      "epoch": 2.209609025481424,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004506585158901422,
      "loss": 5.0701,
      "mean_token_accuracy": 0.20142721831798555,
      "num_tokens": 49264735.0,
      "step": 28400
    },
    {
      "entropy": 5.764660835266113,
      "epoch": 2.2099980548531413,
      "grad_norm": 1.25,
      "learning_rate": 0.0004506411950768207,
      "loss": 5.1332,
      "mean_token_accuracy": 0.19726962000131607,
      "num_tokens": 49273370.0,
      "step": 28405
    },
    {
      "entropy": 5.71741886138916,
      "epoch": 2.210387084224859,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045062387159842363,
      "loss": 5.0665,
      "mean_token_accuracy": 0.2020154267549515,
      "num_tokens": 49282850.0,
      "step": 28410
    },
    {
      "entropy": 5.696995258331299,
      "epoch": 2.2107761135965767,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045060654545521386,
      "loss": 4.987,
      "mean_token_accuracy": 0.2096451461315155,
      "num_tokens": 49290546.0,
      "step": 28415
    },
    {
      "entropy": 5.692635536193848,
      "epoch": 2.211165142968294,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045058921664745436,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2018825277686119,
      "num_tokens": 49298648.0,
      "step": 28420
    },
    {
      "entropy": 5.69274115562439,
      "epoch": 2.2115541723400116,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045057188517540783,
      "loss": 5.0097,
      "mean_token_accuracy": 0.20336603969335557,
      "num_tokens": 49307134.0,
      "step": 28425
    },
    {
      "entropy": 5.720450830459595,
      "epoch": 2.2119432017117293,
      "grad_norm": 1.25,
      "learning_rate": 0.0004505545510393374,
      "loss": 4.9818,
      "mean_token_accuracy": 0.1968682363629341,
      "num_tokens": 49315617.0,
      "step": 28430
    },
    {
      "entropy": 5.727205801010132,
      "epoch": 2.2123322310834466,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045053721423950617,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20569151788949966,
      "num_tokens": 49324435.0,
      "step": 28435
    },
    {
      "entropy": 5.656039524078369,
      "epoch": 2.2127212604551643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004505198747761771,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2061510995030403,
      "num_tokens": 49333758.0,
      "step": 28440
    },
    {
      "entropy": 5.6401002407073975,
      "epoch": 2.213110289826882,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004505025326496132,
      "loss": 4.861,
      "mean_token_accuracy": 0.2155044123530388,
      "num_tokens": 49342643.0,
      "step": 28445
    },
    {
      "entropy": 5.6884276390075685,
      "epoch": 2.2134993191985997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045048518786007784,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21126287281513215,
      "num_tokens": 49351009.0,
      "step": 28450
    },
    {
      "entropy": 5.691434526443482,
      "epoch": 2.213888348570317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045046784040783395,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20124336779117585,
      "num_tokens": 49359415.0,
      "step": 28455
    },
    {
      "entropy": 5.743691778182983,
      "epoch": 2.2142773779420346,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000450450490293145,
      "loss": 5.0952,
      "mean_token_accuracy": 0.19594018459320067,
      "num_tokens": 49367898.0,
      "step": 28460
    },
    {
      "entropy": 5.756645393371582,
      "epoch": 2.2146664073137523,
      "grad_norm": 1.1875,
      "learning_rate": 0.000450433137516274,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20294147729873657,
      "num_tokens": 49376651.0,
      "step": 28465
    },
    {
      "entropy": 5.659844493865966,
      "epoch": 2.2150554366854696,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004504157820774844,
      "loss": 4.9494,
      "mean_token_accuracy": 0.20997174978256225,
      "num_tokens": 49385259.0,
      "step": 28470
    },
    {
      "entropy": 5.702319622039795,
      "epoch": 2.2154444660571873,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004503984239770396,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21031220257282257,
      "num_tokens": 49393868.0,
      "step": 28475
    },
    {
      "entropy": 5.727973079681396,
      "epoch": 2.215833495428905,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004503810632152028,
      "loss": 5.0174,
      "mean_token_accuracy": 0.20674825310707093,
      "num_tokens": 49402062.0,
      "step": 28480
    },
    {
      "entropy": 5.762622737884522,
      "epoch": 2.2162225248006227,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004503636997922375,
      "loss": 5.1105,
      "mean_token_accuracy": 0.19705789387226105,
      "num_tokens": 49411231.0,
      "step": 28485
    },
    {
      "entropy": 5.78741044998169,
      "epoch": 2.21661155417234,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004503463337084072,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2011246770620346,
      "num_tokens": 49419410.0,
      "step": 28490
    },
    {
      "entropy": 5.737607669830322,
      "epoch": 2.2170005835440576,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004503289649639754,
      "loss": 4.9461,
      "mean_token_accuracy": 0.20614745020866393,
      "num_tokens": 49427365.0,
      "step": 28495
    },
    {
      "entropy": 5.709578609466552,
      "epoch": 2.2173896129157753,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004503115935592056,
      "loss": 4.9784,
      "mean_token_accuracy": 0.19668889939785003,
      "num_tokens": 49436395.0,
      "step": 28500
    },
    {
      "entropy": 5.795256042480469,
      "epoch": 2.2177786422874926,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004502942194943614,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20432429760694504,
      "num_tokens": 49446021.0,
      "step": 28505
    },
    {
      "entropy": 5.787047719955444,
      "epoch": 2.2181676716592102,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045027684276970645,
      "loss": 5.1107,
      "mean_token_accuracy": 0.1927965208888054,
      "num_tokens": 49454306.0,
      "step": 28510
    },
    {
      "entropy": 5.712782335281372,
      "epoch": 2.218556701030928,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045025946338550434,
      "loss": 5.0323,
      "mean_token_accuracy": 0.200459261238575,
      "num_tokens": 49462735.0,
      "step": 28515
    },
    {
      "entropy": 5.74784893989563,
      "epoch": 2.218945730402645,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004502420813420189,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19726840108633042,
      "num_tokens": 49471156.0,
      "step": 28520
    },
    {
      "entropy": 5.750462818145752,
      "epoch": 2.219334759774363,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004502246966395137,
      "loss": 5.072,
      "mean_token_accuracy": 0.19947852045297623,
      "num_tokens": 49479505.0,
      "step": 28525
    },
    {
      "entropy": 5.78021764755249,
      "epoch": 2.2197237891460806,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045020730927825274,
      "loss": 5.063,
      "mean_token_accuracy": 0.20308801382780076,
      "num_tokens": 49487879.0,
      "step": 28530
    },
    {
      "entropy": 5.70673394203186,
      "epoch": 2.2201128185177983,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004501899192584997,
      "loss": 4.9127,
      "mean_token_accuracy": 0.20914604663848876,
      "num_tokens": 49496321.0,
      "step": 28535
    },
    {
      "entropy": 5.8107137203216555,
      "epoch": 2.2205018478895155,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004501725265805185,
      "loss": 5.2155,
      "mean_token_accuracy": 0.19303111135959625,
      "num_tokens": 49504955.0,
      "step": 28540
    },
    {
      "entropy": 5.70295205116272,
      "epoch": 2.2208908772612332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045015513124457303,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2039865717291832,
      "num_tokens": 49513678.0,
      "step": 28545
    },
    {
      "entropy": 5.742700242996216,
      "epoch": 2.221279906632951,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004501377332509272,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2036922588944435,
      "num_tokens": 49522315.0,
      "step": 28550
    },
    {
      "entropy": 5.71021728515625,
      "epoch": 2.221668936004668,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045012033259984494,
      "loss": 5.021,
      "mean_token_accuracy": 0.20052336454391478,
      "num_tokens": 49531077.0,
      "step": 28555
    },
    {
      "entropy": 5.727665424346924,
      "epoch": 2.222057965376386,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004501029292915905,
      "loss": 5.0546,
      "mean_token_accuracy": 0.19091786295175553,
      "num_tokens": 49540327.0,
      "step": 28560
    },
    {
      "entropy": 5.733423042297363,
      "epoch": 2.2224469947481036,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045008552332642774,
      "loss": 5.0187,
      "mean_token_accuracy": 0.2003881111741066,
      "num_tokens": 49549440.0,
      "step": 28565
    },
    {
      "entropy": 5.7318004131317135,
      "epoch": 2.222836024119821,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045006811470462083,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2043348714709282,
      "num_tokens": 49558066.0,
      "step": 28570
    },
    {
      "entropy": 5.712904071807861,
      "epoch": 2.2232250534915385,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000450050703426434,
      "loss": 4.9837,
      "mean_token_accuracy": 0.205425925552845,
      "num_tokens": 49566970.0,
      "step": 28575
    },
    {
      "entropy": 5.62446551322937,
      "epoch": 2.2236140828632562,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004500332894921313,
      "loss": 4.8813,
      "mean_token_accuracy": 0.21471214890480042,
      "num_tokens": 49575524.0,
      "step": 28580
    },
    {
      "entropy": 5.764256477355957,
      "epoch": 2.224003112234974,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004500158729019771,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19193635433912276,
      "num_tokens": 49584040.0,
      "step": 28585
    },
    {
      "entropy": 5.765584421157837,
      "epoch": 2.224392141606691,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004499984536562355,
      "loss": 5.0672,
      "mean_token_accuracy": 0.19690656960010527,
      "num_tokens": 49592991.0,
      "step": 28590
    },
    {
      "entropy": 5.726421928405761,
      "epoch": 2.224781170978409,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000449981031755171,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20350545048713684,
      "num_tokens": 49601958.0,
      "step": 28595
    },
    {
      "entropy": 5.681412601470948,
      "epoch": 2.2251702003501266,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044996360719904777,
      "loss": 4.9507,
      "mean_token_accuracy": 0.20510725677013397,
      "num_tokens": 49610624.0,
      "step": 28600
    },
    {
      "entropy": 5.659785270690918,
      "epoch": 2.225559229721844,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004499461799881303,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20861911922693252,
      "num_tokens": 49619308.0,
      "step": 28605
    },
    {
      "entropy": 5.697969436645508,
      "epoch": 2.2259482590935615,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004499287501226831,
      "loss": 4.9826,
      "mean_token_accuracy": 0.19852143824100493,
      "num_tokens": 49627842.0,
      "step": 28610
    },
    {
      "entropy": 5.734337043762207,
      "epoch": 2.2263372884652792,
      "grad_norm": 1.25,
      "learning_rate": 0.00044991131760297045,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1936587244272232,
      "num_tokens": 49636712.0,
      "step": 28615
    },
    {
      "entropy": 5.7494683265686035,
      "epoch": 2.226726317836997,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044989388242925693,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20996786504983903,
      "num_tokens": 49645324.0,
      "step": 28620
    },
    {
      "entropy": 5.698002767562866,
      "epoch": 2.227115347208714,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004498764446018073,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20260098427534104,
      "num_tokens": 49654203.0,
      "step": 28625
    },
    {
      "entropy": 5.77195177078247,
      "epoch": 2.227504376580432,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004498590041208857,
      "loss": 5.0356,
      "mean_token_accuracy": 0.19665776938199997,
      "num_tokens": 49663197.0,
      "step": 28630
    },
    {
      "entropy": 5.688531303405762,
      "epoch": 2.2278934059521496,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004498415609867572,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20388445556163787,
      "num_tokens": 49671993.0,
      "step": 28635
    },
    {
      "entropy": 5.639495372772217,
      "epoch": 2.228282435323867,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044982411519968625,
      "loss": 4.8693,
      "mean_token_accuracy": 0.20722759664058685,
      "num_tokens": 49680403.0,
      "step": 28640
    },
    {
      "entropy": 5.617615652084351,
      "epoch": 2.2286714646955845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004498066667599377,
      "loss": 4.8855,
      "mean_token_accuracy": 0.20946581065654754,
      "num_tokens": 49688667.0,
      "step": 28645
    },
    {
      "entropy": 5.631900691986084,
      "epoch": 2.2290604940673022,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044978921566777604,
      "loss": 5.0217,
      "mean_token_accuracy": 0.19844185262918473,
      "num_tokens": 49696753.0,
      "step": 28650
    },
    {
      "entropy": 5.713625526428222,
      "epoch": 2.2294495234390195,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004497717619234664,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2020262748003006,
      "num_tokens": 49705222.0,
      "step": 28655
    },
    {
      "entropy": 5.72032527923584,
      "epoch": 2.229838552810737,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044975430552727337,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20385972410440445,
      "num_tokens": 49714133.0,
      "step": 28660
    },
    {
      "entropy": 5.65473484992981,
      "epoch": 2.230227582182455,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004497368464794619,
      "loss": 4.9228,
      "mean_token_accuracy": 0.20938296765089034,
      "num_tokens": 49722743.0,
      "step": 28665
    },
    {
      "entropy": 5.672687244415283,
      "epoch": 2.230616611554172,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044971938478029693,
      "loss": 4.897,
      "mean_token_accuracy": 0.21367636770009996,
      "num_tokens": 49731721.0,
      "step": 28670
    },
    {
      "entropy": 5.661153173446655,
      "epoch": 2.23100564092589,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044970192043004343,
      "loss": 4.9059,
      "mean_token_accuracy": 0.20490513890981674,
      "num_tokens": 49740373.0,
      "step": 28675
    },
    {
      "entropy": 5.67064847946167,
      "epoch": 2.2313946702976075,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004496844534289663,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20849578976631164,
      "num_tokens": 49748544.0,
      "step": 28680
    },
    {
      "entropy": 5.7935545444488525,
      "epoch": 2.231783699669325,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044966698377733064,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20090939104557037,
      "num_tokens": 49757738.0,
      "step": 28685
    },
    {
      "entropy": 5.670059394836426,
      "epoch": 2.2321727290410425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004496495114754015,
      "loss": 4.9064,
      "mean_token_accuracy": 0.20539970695972443,
      "num_tokens": 49766203.0,
      "step": 28690
    },
    {
      "entropy": 5.663131141662598,
      "epoch": 2.23256175841276,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044963203652344404,
      "loss": 4.9676,
      "mean_token_accuracy": 0.20211831033229827,
      "num_tokens": 49774951.0,
      "step": 28695
    },
    {
      "entropy": 5.610601711273193,
      "epoch": 2.232950787784478,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004496145589217234,
      "loss": 4.9732,
      "mean_token_accuracy": 0.2016509145498276,
      "num_tokens": 49783250.0,
      "step": 28700
    },
    {
      "entropy": 5.667260837554932,
      "epoch": 2.233339817156195,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044959707867050467,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2071501135826111,
      "num_tokens": 49792832.0,
      "step": 28705
    },
    {
      "entropy": 5.713402366638183,
      "epoch": 2.233728846527913,
      "grad_norm": 1.25,
      "learning_rate": 0.0004495795957700532,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21105409860610963,
      "num_tokens": 49800745.0,
      "step": 28710
    },
    {
      "entropy": 5.618764972686767,
      "epoch": 2.2341178758996305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044956211022063424,
      "loss": 4.886,
      "mean_token_accuracy": 0.21803490072488785,
      "num_tokens": 49809458.0,
      "step": 28715
    },
    {
      "entropy": 5.70646824836731,
      "epoch": 2.234506905271348,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044954462202251316,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2043460339307785,
      "num_tokens": 49818122.0,
      "step": 28720
    },
    {
      "entropy": 5.653728866577149,
      "epoch": 2.2348959346430655,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044952713117595516,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2052289515733719,
      "num_tokens": 49826451.0,
      "step": 28725
    },
    {
      "entropy": 5.785553741455078,
      "epoch": 2.235284964014783,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044950963768122576,
      "loss": 5.0224,
      "mean_token_accuracy": 0.19605523645877837,
      "num_tokens": 49835463.0,
      "step": 28730
    },
    {
      "entropy": 5.71084156036377,
      "epoch": 2.235673993386501,
      "grad_norm": 1.25,
      "learning_rate": 0.0004494921415385905,
      "loss": 4.966,
      "mean_token_accuracy": 0.20406141579151155,
      "num_tokens": 49843896.0,
      "step": 28735
    },
    {
      "entropy": 5.738232374191284,
      "epoch": 2.236063022758218,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044947464274831464,
      "loss": 5.0074,
      "mean_token_accuracy": 0.19478444308042525,
      "num_tokens": 49852380.0,
      "step": 28740
    },
    {
      "entropy": 5.766475057601928,
      "epoch": 2.236452052129936,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004494571413106637,
      "loss": 4.9561,
      "mean_token_accuracy": 0.20915485322475433,
      "num_tokens": 49861290.0,
      "step": 28745
    },
    {
      "entropy": 5.779478549957275,
      "epoch": 2.2368410815016535,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044943963722590344,
      "loss": 5.0126,
      "mean_token_accuracy": 0.19861364364624023,
      "num_tokens": 49869295.0,
      "step": 28750
    },
    {
      "entropy": 5.694930696487427,
      "epoch": 2.2372301108733708,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044942213049429917,
      "loss": 5.0783,
      "mean_token_accuracy": 0.19997182488441467,
      "num_tokens": 49877771.0,
      "step": 28755
    },
    {
      "entropy": 5.717388725280761,
      "epoch": 2.2376191402450885,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044940462111611685,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20335056036710739,
      "num_tokens": 49886531.0,
      "step": 28760
    },
    {
      "entropy": 5.754183197021485,
      "epoch": 2.238008169616806,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044938710909162195,
      "loss": 5.0634,
      "mean_token_accuracy": 0.2055935740470886,
      "num_tokens": 49895323.0,
      "step": 28765
    },
    {
      "entropy": 5.770350980758667,
      "epoch": 2.2383971989885234,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004493695944210802,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20306635349988938,
      "num_tokens": 49903918.0,
      "step": 28770
    },
    {
      "entropy": 5.680583381652832,
      "epoch": 2.238786228360241,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004493520771047574,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2070600837469101,
      "num_tokens": 49912754.0,
      "step": 28775
    },
    {
      "entropy": 5.7315887928009035,
      "epoch": 2.239175257731959,
      "grad_norm": 1.125,
      "learning_rate": 0.00044933455714291944,
      "loss": 5.0617,
      "mean_token_accuracy": 0.20385382771492006,
      "num_tokens": 49922188.0,
      "step": 28780
    },
    {
      "entropy": 5.744244050979614,
      "epoch": 2.2395642871036765,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044931703453583195,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20337348729372023,
      "num_tokens": 49930718.0,
      "step": 28785
    },
    {
      "entropy": 5.700878524780274,
      "epoch": 2.2399533164753938,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044929950928376095,
      "loss": 5.0645,
      "mean_token_accuracy": 0.20703099966049193,
      "num_tokens": 49939095.0,
      "step": 28790
    },
    {
      "entropy": 5.724122047424316,
      "epoch": 2.2403423458471114,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004492819813869723,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20994768142700196,
      "num_tokens": 49946996.0,
      "step": 28795
    },
    {
      "entropy": 5.742649364471435,
      "epoch": 2.240731375218829,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044926445084573206,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20634025633335112,
      "num_tokens": 49956262.0,
      "step": 28800
    },
    {
      "entropy": 5.70239667892456,
      "epoch": 2.2411204045905464,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000449246917660306,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20563437938690185,
      "num_tokens": 49964581.0,
      "step": 28805
    },
    {
      "entropy": 5.578705787658691,
      "epoch": 2.241509433962264,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004492293818309604,
      "loss": 4.871,
      "mean_token_accuracy": 0.21165210902690887,
      "num_tokens": 49972876.0,
      "step": 28810
    },
    {
      "entropy": 5.700698709487915,
      "epoch": 2.241898463333982,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004492118433579613,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21003865748643874,
      "num_tokens": 49981605.0,
      "step": 28815
    },
    {
      "entropy": 5.803601694107056,
      "epoch": 2.2422874927056995,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044919430224157463,
      "loss": 5.1345,
      "mean_token_accuracy": 0.19914266020059584,
      "num_tokens": 49990576.0,
      "step": 28820
    },
    {
      "entropy": 5.726606035232544,
      "epoch": 2.2426765220774167,
      "grad_norm": 1.25,
      "learning_rate": 0.00044917675848206684,
      "loss": 4.9562,
      "mean_token_accuracy": 0.20665538758039476,
      "num_tokens": 49999350.0,
      "step": 28825
    },
    {
      "entropy": 5.64574031829834,
      "epoch": 2.2430655514491344,
      "grad_norm": 1.125,
      "learning_rate": 0.00044915921207970387,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20463927090168,
      "num_tokens": 50008202.0,
      "step": 28830
    },
    {
      "entropy": 5.670968437194825,
      "epoch": 2.243454580820852,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044914166303475206,
      "loss": 4.8907,
      "mean_token_accuracy": 0.2097530260682106,
      "num_tokens": 50016600.0,
      "step": 28835
    },
    {
      "entropy": 5.716279983520508,
      "epoch": 2.2438436101925694,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044912411134747764,
      "loss": 4.9291,
      "mean_token_accuracy": 0.20768495351076127,
      "num_tokens": 50025029.0,
      "step": 28840
    },
    {
      "entropy": 5.618334722518921,
      "epoch": 2.244232639564287,
      "grad_norm": 1.25,
      "learning_rate": 0.00044910655701814705,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2046564906835556,
      "num_tokens": 50033692.0,
      "step": 28845
    },
    {
      "entropy": 5.725336933135987,
      "epoch": 2.244621668936005,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004490890000470266,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2019438162446022,
      "num_tokens": 50042572.0,
      "step": 28850
    },
    {
      "entropy": 5.801209402084351,
      "epoch": 2.245010698307722,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044907144043438265,
      "loss": 5.079,
      "mean_token_accuracy": 0.1992827132344246,
      "num_tokens": 50050823.0,
      "step": 28855
    },
    {
      "entropy": 5.721604251861573,
      "epoch": 2.2453997276794397,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004490538781804816,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20633311569690704,
      "num_tokens": 50058596.0,
      "step": 28860
    },
    {
      "entropy": 5.6724766254425045,
      "epoch": 2.2457887570511574,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044903631328559,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21207424402236938,
      "num_tokens": 50067482.0,
      "step": 28865
    },
    {
      "entropy": 5.693482780456543,
      "epoch": 2.246177786422875,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044901874574997444,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20299184322357178,
      "num_tokens": 50075581.0,
      "step": 28870
    },
    {
      "entropy": 5.688263368606568,
      "epoch": 2.2465668157945924,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044900117557390137,
      "loss": 4.927,
      "mean_token_accuracy": 0.21165452003479004,
      "num_tokens": 50083906.0,
      "step": 28875
    },
    {
      "entropy": 5.7334206104278564,
      "epoch": 2.24695584516631,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044898360275763733,
      "loss": 5.029,
      "mean_token_accuracy": 0.19959970265626908,
      "num_tokens": 50091494.0,
      "step": 28880
    },
    {
      "entropy": 5.72422685623169,
      "epoch": 2.247344874538028,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044896602730144905,
      "loss": 5.0652,
      "mean_token_accuracy": 0.20665520280599595,
      "num_tokens": 50100576.0,
      "step": 28885
    },
    {
      "entropy": 5.781793928146362,
      "epoch": 2.247733903909745,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044894844920560323,
      "loss": 5.1414,
      "mean_token_accuracy": 0.18653077632188797,
      "num_tokens": 50109832.0,
      "step": 28890
    },
    {
      "entropy": 5.7336241722106935,
      "epoch": 2.2481229332814627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004489308684703666,
      "loss": 5.114,
      "mean_token_accuracy": 0.1952989712357521,
      "num_tokens": 50120038.0,
      "step": 28895
    },
    {
      "entropy": 5.707908487319946,
      "epoch": 2.2485119626531804,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004489132850960059,
      "loss": 5.0391,
      "mean_token_accuracy": 0.2019359603524208,
      "num_tokens": 50128580.0,
      "step": 28900
    },
    {
      "entropy": 5.746394538879395,
      "epoch": 2.2489009920248977,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004488956990827878,
      "loss": 5.0867,
      "mean_token_accuracy": 0.19932993948459626,
      "num_tokens": 50137540.0,
      "step": 28905
    },
    {
      "entropy": 5.764518928527832,
      "epoch": 2.2492900213966154,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004488781104309793,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20121348202228545,
      "num_tokens": 50146292.0,
      "step": 28910
    },
    {
      "entropy": 5.698071718215942,
      "epoch": 2.249679050768333,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044886051914084726,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2128889575600624,
      "num_tokens": 50154893.0,
      "step": 28915
    },
    {
      "entropy": 5.749345064163208,
      "epoch": 2.2500680801400508,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044884292521265846,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20161522775888444,
      "num_tokens": 50163707.0,
      "step": 28920
    },
    {
      "entropy": 5.736520004272461,
      "epoch": 2.250457109511768,
      "grad_norm": 1.125,
      "learning_rate": 0.0004488253286466801,
      "loss": 4.9818,
      "mean_token_accuracy": 0.20222944170236587,
      "num_tokens": 50172062.0,
      "step": 28925
    },
    {
      "entropy": 5.69788556098938,
      "epoch": 2.2508461388834857,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044880772944317905,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20335436165332793,
      "num_tokens": 50180362.0,
      "step": 28930
    },
    {
      "entropy": 5.734890508651733,
      "epoch": 2.2512351682552034,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044879012760242224,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20479643642902373,
      "num_tokens": 50189038.0,
      "step": 28935
    },
    {
      "entropy": 5.764187574386597,
      "epoch": 2.2516241976269207,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044877252312467694,
      "loss": 4.9696,
      "mean_token_accuracy": 0.20019122958183289,
      "num_tokens": 50198288.0,
      "step": 28940
    },
    {
      "entropy": 5.643303442001343,
      "epoch": 2.2520132269986384,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004487549160102101,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2081664353609085,
      "num_tokens": 50207442.0,
      "step": 28945
    },
    {
      "entropy": 5.728520298004151,
      "epoch": 2.252402256370356,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044873730625928914,
      "loss": 5.0732,
      "mean_token_accuracy": 0.19978389739990235,
      "num_tokens": 50216794.0,
      "step": 28950
    },
    {
      "entropy": 5.770139837265015,
      "epoch": 2.2527912857420738,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044871969387218094,
      "loss": 4.9436,
      "mean_token_accuracy": 0.20935565531253814,
      "num_tokens": 50224764.0,
      "step": 28955
    },
    {
      "entropy": 5.712708139419556,
      "epoch": 2.253180315113791,
      "grad_norm": 1.234375,
      "learning_rate": 0.000448702078849153,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19823368936777114,
      "num_tokens": 50233256.0,
      "step": 28960
    },
    {
      "entropy": 5.715312337875366,
      "epoch": 2.2535693444855087,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044868446119047234,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2066602662205696,
      "num_tokens": 50242862.0,
      "step": 28965
    },
    {
      "entropy": 5.705289602279663,
      "epoch": 2.253958373857226,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004486668408964065,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20061289072036742,
      "num_tokens": 50252048.0,
      "step": 28970
    },
    {
      "entropy": 5.772410678863525,
      "epoch": 2.2543474032289437,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004486492179672228,
      "loss": 5.0801,
      "mean_token_accuracy": 0.20077840089797974,
      "num_tokens": 50260545.0,
      "step": 28975
    },
    {
      "entropy": 5.707185697555542,
      "epoch": 2.2547364326006614,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044863159240318863,
      "loss": 4.9246,
      "mean_token_accuracy": 0.20922300964593887,
      "num_tokens": 50268714.0,
      "step": 28980
    },
    {
      "entropy": 5.653179931640625,
      "epoch": 2.255125461972379,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004486139642045714,
      "loss": 4.9012,
      "mean_token_accuracy": 0.21264299750328064,
      "num_tokens": 50276876.0,
      "step": 28985
    },
    {
      "entropy": 5.705392980575562,
      "epoch": 2.2555144913440963,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004485963333716385,
      "loss": 5.0708,
      "mean_token_accuracy": 0.19574485272169112,
      "num_tokens": 50285924.0,
      "step": 28990
    },
    {
      "entropy": 5.752621698379516,
      "epoch": 2.255903520715814,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004485786999046576,
      "loss": 5.0299,
      "mean_token_accuracy": 0.19410956054925918,
      "num_tokens": 50295358.0,
      "step": 28995
    },
    {
      "entropy": 5.752086639404297,
      "epoch": 2.2562925500875317,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044856106380389624,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20081109553575516,
      "num_tokens": 50303572.0,
      "step": 29000
    },
    {
      "entropy": 5.7104905128479,
      "epoch": 2.256681579459249,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000448543425069622,
      "loss": 4.9541,
      "mean_token_accuracy": 0.21394867449998856,
      "num_tokens": 50312533.0,
      "step": 29005
    },
    {
      "entropy": 5.677719306945801,
      "epoch": 2.2570706088309667,
      "grad_norm": 1.25,
      "learning_rate": 0.0004485257837021025,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2056881830096245,
      "num_tokens": 50321049.0,
      "step": 29010
    },
    {
      "entropy": 5.652480173110962,
      "epoch": 2.2574596382026844,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044850813970160534,
      "loss": 4.919,
      "mean_token_accuracy": 0.21142540723085404,
      "num_tokens": 50329308.0,
      "step": 29015
    },
    {
      "entropy": 5.7265726089477536,
      "epoch": 2.257848667574402,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004484904930683984,
      "loss": 4.9418,
      "mean_token_accuracy": 0.2110203981399536,
      "num_tokens": 50337311.0,
      "step": 29020
    },
    {
      "entropy": 5.622326374053955,
      "epoch": 2.2582376969461193,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004484728438027494,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2086610645055771,
      "num_tokens": 50346557.0,
      "step": 29025
    },
    {
      "entropy": 5.619020271301269,
      "epoch": 2.258626726317837,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004484551919049261,
      "loss": 4.959,
      "mean_token_accuracy": 0.21072124242782592,
      "num_tokens": 50354982.0,
      "step": 29030
    },
    {
      "entropy": 5.698486852645874,
      "epoch": 2.2590157556895547,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044843753737519625,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20360653698444367,
      "num_tokens": 50363414.0,
      "step": 29035
    },
    {
      "entropy": 5.73218789100647,
      "epoch": 2.259404785061272,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004484198802138279,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2033605694770813,
      "num_tokens": 50371803.0,
      "step": 29040
    },
    {
      "entropy": 5.747698450088501,
      "epoch": 2.2597938144329897,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004484022204210889,
      "loss": 5.027,
      "mean_token_accuracy": 0.2098245769739151,
      "num_tokens": 50380125.0,
      "step": 29045
    },
    {
      "entropy": 5.622207450866699,
      "epoch": 2.2601828438047074,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044838455799724717,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2039381295442581,
      "num_tokens": 50388916.0,
      "step": 29050
    },
    {
      "entropy": 5.7461292266845705,
      "epoch": 2.260571873176425,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044836689294257075,
      "loss": 5.0642,
      "mean_token_accuracy": 0.2043704330921173,
      "num_tokens": 50397353.0,
      "step": 29055
    },
    {
      "entropy": 5.709478807449341,
      "epoch": 2.2609609025481423,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004483492252573276,
      "loss": 4.9679,
      "mean_token_accuracy": 0.21059502512216569,
      "num_tokens": 50405176.0,
      "step": 29060
    },
    {
      "entropy": 5.649062013626098,
      "epoch": 2.26134993191986,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000448331554941786,
      "loss": 4.9836,
      "mean_token_accuracy": 0.20330188274383545,
      "num_tokens": 50413845.0,
      "step": 29065
    },
    {
      "entropy": 5.692983388900757,
      "epoch": 2.2617389612915777,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044831388199621385,
      "loss": 5.0656,
      "mean_token_accuracy": 0.20101059079170228,
      "num_tokens": 50422343.0,
      "step": 29070
    },
    {
      "entropy": 5.670225954055786,
      "epoch": 2.262127990663295,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044829620642087946,
      "loss": 4.9419,
      "mean_token_accuracy": 0.20933328717947006,
      "num_tokens": 50430438.0,
      "step": 29075
    },
    {
      "entropy": 5.680683898925781,
      "epoch": 2.2625170200350126,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004482785282160509,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19926034957170485,
      "num_tokens": 50438865.0,
      "step": 29080
    },
    {
      "entropy": 5.738600778579712,
      "epoch": 2.2629060494067303,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044826084738199657,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20064839571714402,
      "num_tokens": 50447784.0,
      "step": 29085
    },
    {
      "entropy": 5.805469036102295,
      "epoch": 2.2632950787784476,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044824316391898464,
      "loss": 5.0254,
      "mean_token_accuracy": 0.2003640577197075,
      "num_tokens": 50457266.0,
      "step": 29090
    },
    {
      "entropy": 5.817899703979492,
      "epoch": 2.2636841081501653,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044822547782728334,
      "loss": 5.1791,
      "mean_token_accuracy": 0.19159330576658248,
      "num_tokens": 50466171.0,
      "step": 29095
    },
    {
      "entropy": 5.750521087646485,
      "epoch": 2.264073137521883,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004482077891071612,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2073726922273636,
      "num_tokens": 50474127.0,
      "step": 29100
    },
    {
      "entropy": 5.695936059951782,
      "epoch": 2.2644621668936002,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044819009775888655,
      "loss": 4.9262,
      "mean_token_accuracy": 0.20355207473039627,
      "num_tokens": 50482724.0,
      "step": 29105
    },
    {
      "entropy": 5.820371866226196,
      "epoch": 2.264851196265318,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044817240378272784,
      "loss": 5.1183,
      "mean_token_accuracy": 0.19183087348937988,
      "num_tokens": 50491412.0,
      "step": 29110
    },
    {
      "entropy": 5.79158124923706,
      "epoch": 2.2652402256370356,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044815470717895345,
      "loss": 5.1236,
      "mean_token_accuracy": 0.19906512051820754,
      "num_tokens": 50499843.0,
      "step": 29115
    },
    {
      "entropy": 5.750116157531738,
      "epoch": 2.2656292550087533,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004481370079478321,
      "loss": 5.0446,
      "mean_token_accuracy": 0.20204756557941436,
      "num_tokens": 50509501.0,
      "step": 29120
    },
    {
      "entropy": 5.773785543441773,
      "epoch": 2.2660182843804706,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004481193060896322,
      "loss": 5.025,
      "mean_token_accuracy": 0.2096179336309433,
      "num_tokens": 50518438.0,
      "step": 29125
    },
    {
      "entropy": 5.763443231582642,
      "epoch": 2.2664073137521883,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004481016016046223,
      "loss": 5.0106,
      "mean_token_accuracy": 0.20079202353954315,
      "num_tokens": 50527516.0,
      "step": 29130
    },
    {
      "entropy": 5.693717288970947,
      "epoch": 2.266796343123906,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004480838944930712,
      "loss": 4.9563,
      "mean_token_accuracy": 0.20296838879585266,
      "num_tokens": 50536179.0,
      "step": 29135
    },
    {
      "entropy": 5.7756524085998535,
      "epoch": 2.2671853724956232,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004480661847552474,
      "loss": 5.0152,
      "mean_token_accuracy": 0.20065454244613648,
      "num_tokens": 50545231.0,
      "step": 29140
    },
    {
      "entropy": 5.726790904998779,
      "epoch": 2.267574401867341,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004480484723914198,
      "loss": 5.0197,
      "mean_token_accuracy": 0.20391011983156204,
      "num_tokens": 50553873.0,
      "step": 29145
    },
    {
      "entropy": 5.7086320400238035,
      "epoch": 2.2679634312390586,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004480307574018571,
      "loss": 5.0209,
      "mean_token_accuracy": 0.2058439403772354,
      "num_tokens": 50563733.0,
      "step": 29150
    },
    {
      "entropy": 5.722029399871826,
      "epoch": 2.2683524606107763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000448013039786828,
      "loss": 4.9598,
      "mean_token_accuracy": 0.20372679829597473,
      "num_tokens": 50573187.0,
      "step": 29155
    },
    {
      "entropy": 5.7034180641174315,
      "epoch": 2.2687414899824936,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044799531954660133,
      "loss": 4.9563,
      "mean_token_accuracy": 0.20959169864654542,
      "num_tokens": 50581580.0,
      "step": 29160
    },
    {
      "entropy": 5.731806850433349,
      "epoch": 2.2691305193542113,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004479775966814461,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20950349867343904,
      "num_tokens": 50589471.0,
      "step": 29165
    },
    {
      "entropy": 5.727677726745606,
      "epoch": 2.269519548725929,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044795987119163115,
      "loss": 5.013,
      "mean_token_accuracy": 0.20892661958932876,
      "num_tokens": 50597923.0,
      "step": 29170
    },
    {
      "entropy": 5.8044098854064945,
      "epoch": 2.2699085780976462,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004479421430774255,
      "loss": 5.1011,
      "mean_token_accuracy": 0.189509841799736,
      "num_tokens": 50606981.0,
      "step": 29175
    },
    {
      "entropy": 5.755601930618286,
      "epoch": 2.270297607469364,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004479244123390981,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20328839421272277,
      "num_tokens": 50615680.0,
      "step": 29180
    },
    {
      "entropy": 5.767923974990845,
      "epoch": 2.2706866368410816,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044790667897691796,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2092898041009903,
      "num_tokens": 50624053.0,
      "step": 29185
    },
    {
      "entropy": 5.713153457641601,
      "epoch": 2.271075666212799,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044788894299115415,
      "loss": 5.006,
      "mean_token_accuracy": 0.20494316667318344,
      "num_tokens": 50632671.0,
      "step": 29190
    },
    {
      "entropy": 5.723009634017944,
      "epoch": 2.2714646955845166,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004478712043820758,
      "loss": 4.9874,
      "mean_token_accuracy": 0.20882837772369384,
      "num_tokens": 50641734.0,
      "step": 29195
    },
    {
      "entropy": 5.704870271682739,
      "epoch": 2.2718537249562343,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004478534631499521,
      "loss": 4.9489,
      "mean_token_accuracy": 0.20966098308563233,
      "num_tokens": 50650711.0,
      "step": 29200
    },
    {
      "entropy": 5.737490129470825,
      "epoch": 2.2722427543279515,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044783571929505223,
      "loss": 5.0156,
      "mean_token_accuracy": 0.1974651038646698,
      "num_tokens": 50659814.0,
      "step": 29205
    },
    {
      "entropy": 5.699716663360595,
      "epoch": 2.2726317836996692,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004478179728176454,
      "loss": 4.976,
      "mean_token_accuracy": 0.20993762165308,
      "num_tokens": 50668340.0,
      "step": 29210
    },
    {
      "entropy": 5.697183990478516,
      "epoch": 2.273020813071387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004478002237180009,
      "loss": 4.944,
      "mean_token_accuracy": 0.20827334970235825,
      "num_tokens": 50676847.0,
      "step": 29215
    },
    {
      "entropy": 5.623721075057984,
      "epoch": 2.2734098424431046,
      "grad_norm": 1.25,
      "learning_rate": 0.000447782471996388,
      "loss": 4.979,
      "mean_token_accuracy": 0.20565496683120726,
      "num_tokens": 50685074.0,
      "step": 29220
    },
    {
      "entropy": 5.721696615219116,
      "epoch": 2.273798871814822,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004477647176530762,
      "loss": 5.0316,
      "mean_token_accuracy": 0.201720666885376,
      "num_tokens": 50693386.0,
      "step": 29225
    },
    {
      "entropy": 5.712011814117432,
      "epoch": 2.2741879011865396,
      "grad_norm": 1.375,
      "learning_rate": 0.0004477469606883347,
      "loss": 4.977,
      "mean_token_accuracy": 0.20554226338863374,
      "num_tokens": 50701966.0,
      "step": 29230
    },
    {
      "entropy": 5.678589010238648,
      "epoch": 2.2745769305582573,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004477292011024331,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2078375920653343,
      "num_tokens": 50710630.0,
      "step": 29235
    },
    {
      "entropy": 5.675682258605957,
      "epoch": 2.2749659599299745,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044771143889564075,
      "loss": 4.94,
      "mean_token_accuracy": 0.20893006920814514,
      "num_tokens": 50719415.0,
      "step": 29240
    },
    {
      "entropy": 5.687006664276123,
      "epoch": 2.275354989301692,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044769367406822727,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21803331673145293,
      "num_tokens": 50728064.0,
      "step": 29245
    },
    {
      "entropy": 5.669770240783691,
      "epoch": 2.27574401867341,
      "grad_norm": 1.125,
      "learning_rate": 0.0004476759066204621,
      "loss": 4.9875,
      "mean_token_accuracy": 0.203485344350338,
      "num_tokens": 50736978.0,
      "step": 29250
    },
    {
      "entropy": 5.754757022857666,
      "epoch": 2.2761330480451276,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004476581365526149,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2006853401660919,
      "num_tokens": 50745368.0,
      "step": 29255
    },
    {
      "entropy": 5.729412174224853,
      "epoch": 2.276522077416845,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004476403638649553,
      "loss": 4.9541,
      "mean_token_accuracy": 0.21530385911464692,
      "num_tokens": 50753765.0,
      "step": 29260
    },
    {
      "entropy": 5.7657183647155765,
      "epoch": 2.2769111067885626,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044762258855775294,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20042400062084198,
      "num_tokens": 50762165.0,
      "step": 29265
    },
    {
      "entropy": 5.726515674591065,
      "epoch": 2.2773001361602803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044760481063127754,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19992605298757554,
      "num_tokens": 50771837.0,
      "step": 29270
    },
    {
      "entropy": 5.7414594173431395,
      "epoch": 2.2776891655319975,
      "grad_norm": 1.125,
      "learning_rate": 0.000447587030085799,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20236752778291703,
      "num_tokens": 50781518.0,
      "step": 29275
    },
    {
      "entropy": 5.740006732940674,
      "epoch": 2.278078194903715,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044756924692158684,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20719189643859864,
      "num_tokens": 50790708.0,
      "step": 29280
    },
    {
      "entropy": 5.70249605178833,
      "epoch": 2.278467224275433,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004475514611389111,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1995113044977188,
      "num_tokens": 50798783.0,
      "step": 29285
    },
    {
      "entropy": 5.6883961200714115,
      "epoch": 2.2788562536471506,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004475336727380415,
      "loss": 4.927,
      "mean_token_accuracy": 0.20579149574041367,
      "num_tokens": 50807033.0,
      "step": 29290
    },
    {
      "entropy": 5.655897045135498,
      "epoch": 2.279245283018868,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004475158817192481,
      "loss": 4.8726,
      "mean_token_accuracy": 0.20687298625707626,
      "num_tokens": 50815548.0,
      "step": 29295
    },
    {
      "entropy": 5.714279890060425,
      "epoch": 2.2796343123905856,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044749808808280077,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2046896204352379,
      "num_tokens": 50824268.0,
      "step": 29300
    },
    {
      "entropy": 5.679475545883179,
      "epoch": 2.280023341762303,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044748029182896956,
      "loss": 4.903,
      "mean_token_accuracy": 0.2114642381668091,
      "num_tokens": 50833136.0,
      "step": 29305
    },
    {
      "entropy": 5.776253938674927,
      "epoch": 2.2804123711340205,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004474624929580243,
      "loss": 5.0678,
      "mean_token_accuracy": 0.202504725754261,
      "num_tokens": 50841911.0,
      "step": 29310
    },
    {
      "entropy": 5.799944019317627,
      "epoch": 2.280801400505738,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004474446914702354,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19442682564258576,
      "num_tokens": 50850402.0,
      "step": 29315
    },
    {
      "entropy": 5.768575286865234,
      "epoch": 2.281190429877456,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004474268873658727,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20455536097288132,
      "num_tokens": 50858473.0,
      "step": 29320
    },
    {
      "entropy": 5.690868806838989,
      "epoch": 2.281579459249173,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004474090806452064,
      "loss": 4.8735,
      "mean_token_accuracy": 0.21203890442848206,
      "num_tokens": 50866798.0,
      "step": 29325
    },
    {
      "entropy": 5.648712921142578,
      "epoch": 2.281968488620891,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044739127130850675,
      "loss": 4.9508,
      "mean_token_accuracy": 0.20579357147216798,
      "num_tokens": 50875279.0,
      "step": 29330
    },
    {
      "entropy": 5.613885068893433,
      "epoch": 2.2823575179926086,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044737345935604397,
      "loss": 4.8693,
      "mean_token_accuracy": 0.2129216656088829,
      "num_tokens": 50884021.0,
      "step": 29335
    },
    {
      "entropy": 5.640703439712524,
      "epoch": 2.282746547364326,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004473556447880883,
      "loss": 4.951,
      "mean_token_accuracy": 0.20622283816337586,
      "num_tokens": 50891722.0,
      "step": 29340
    },
    {
      "entropy": 5.6629375457763675,
      "epoch": 2.2831355767360435,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044733782760490997,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21149497628211975,
      "num_tokens": 50900985.0,
      "step": 29345
    },
    {
      "entropy": 5.76675853729248,
      "epoch": 2.283524606107761,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004473200078067795,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19177106618881226,
      "num_tokens": 50909750.0,
      "step": 29350
    },
    {
      "entropy": 5.759437704086304,
      "epoch": 2.283913635479479,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044730218539396717,
      "loss": 5.0466,
      "mean_token_accuracy": 0.19571248888969422,
      "num_tokens": 50918563.0,
      "step": 29355
    },
    {
      "entropy": 5.741450548171997,
      "epoch": 2.284302664851196,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004472843603667434,
      "loss": 5.0722,
      "mean_token_accuracy": 0.19488185048103332,
      "num_tokens": 50927503.0,
      "step": 29360
    },
    {
      "entropy": 5.69946608543396,
      "epoch": 2.284691694222914,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044726653272537866,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20893163681030275,
      "num_tokens": 50936198.0,
      "step": 29365
    },
    {
      "entropy": 5.641592311859131,
      "epoch": 2.2850807235946315,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004472487024701435,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2048219621181488,
      "num_tokens": 50945589.0,
      "step": 29370
    },
    {
      "entropy": 5.69799575805664,
      "epoch": 2.285469752966349,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004472308696013085,
      "loss": 4.949,
      "mean_token_accuracy": 0.21066120713949205,
      "num_tokens": 50953524.0,
      "step": 29375
    },
    {
      "entropy": 5.700119113922119,
      "epoch": 2.2858587823380665,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004472130341191441,
      "loss": 5.0144,
      "mean_token_accuracy": 0.19730790406465532,
      "num_tokens": 50962127.0,
      "step": 29380
    },
    {
      "entropy": 5.699433088302612,
      "epoch": 2.286247811709784,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004471951960239211,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20711107105016707,
      "num_tokens": 50970662.0,
      "step": 29385
    },
    {
      "entropy": 5.668705081939697,
      "epoch": 2.286636841081502,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044717735531591,
      "loss": 4.9237,
      "mean_token_accuracy": 0.21477384865283966,
      "num_tokens": 50979182.0,
      "step": 29390
    },
    {
      "entropy": 5.713355875015258,
      "epoch": 2.287025870453219,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044715951199538156,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20238587409257888,
      "num_tokens": 50987285.0,
      "step": 29395
    },
    {
      "entropy": 5.741268301010132,
      "epoch": 2.287414899824937,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044714166606260657,
      "loss": 5.01,
      "mean_token_accuracy": 0.2034016028046608,
      "num_tokens": 50995988.0,
      "step": 29400
    },
    {
      "entropy": 5.6798093795776365,
      "epoch": 2.2878039291966545,
      "grad_norm": 1.375,
      "learning_rate": 0.0004471238175178559,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2053341180086136,
      "num_tokens": 51003963.0,
      "step": 29405
    },
    {
      "entropy": 5.683936405181885,
      "epoch": 2.288192958568372,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004471059663614002,
      "loss": 4.9639,
      "mean_token_accuracy": 0.20018507540225983,
      "num_tokens": 51012429.0,
      "step": 29410
    },
    {
      "entropy": 5.738192701339722,
      "epoch": 2.2885819879400895,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004470881125935103,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20345714092254638,
      "num_tokens": 51021321.0,
      "step": 29415
    },
    {
      "entropy": 5.702590274810791,
      "epoch": 2.288971017311807,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044707025621445735,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2062503829598427,
      "num_tokens": 51029713.0,
      "step": 29420
    },
    {
      "entropy": 5.664555883407592,
      "epoch": 2.2893600466835244,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044705239722451194,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20109730809926987,
      "num_tokens": 51038832.0,
      "step": 29425
    },
    {
      "entropy": 5.714535713195801,
      "epoch": 2.289749076055242,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044703453562394544,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21457421481609346,
      "num_tokens": 51046786.0,
      "step": 29430
    },
    {
      "entropy": 5.700632524490357,
      "epoch": 2.29013810542696,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004470166714130286,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20369420051574708,
      "num_tokens": 51054855.0,
      "step": 29435
    },
    {
      "entropy": 5.769228935241699,
      "epoch": 2.290527134798677,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044699880459203264,
      "loss": 5.1224,
      "mean_token_accuracy": 0.20373376607894897,
      "num_tokens": 51063098.0,
      "step": 29440
    },
    {
      "entropy": 5.680781698226928,
      "epoch": 2.290916164170395,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044698093516122844,
      "loss": 4.8757,
      "mean_token_accuracy": 0.21869378834962844,
      "num_tokens": 51071140.0,
      "step": 29445
    },
    {
      "entropy": 5.707134532928467,
      "epoch": 2.2913051935421125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004469630631208874,
      "loss": 5.001,
      "mean_token_accuracy": 0.20636340528726577,
      "num_tokens": 51079698.0,
      "step": 29450
    },
    {
      "entropy": 5.709994173049926,
      "epoch": 2.29169422291383,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004469451884712805,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20447997003793716,
      "num_tokens": 51088900.0,
      "step": 29455
    },
    {
      "entropy": 5.684857654571533,
      "epoch": 2.2920832522855474,
      "grad_norm": 1.25,
      "learning_rate": 0.0004469273112126791,
      "loss": 4.9225,
      "mean_token_accuracy": 0.20729732662439346,
      "num_tokens": 51098289.0,
      "step": 29460
    },
    {
      "entropy": 5.752219724655151,
      "epoch": 2.292472281657265,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044690943134535445,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2085706725716591,
      "num_tokens": 51106795.0,
      "step": 29465
    },
    {
      "entropy": 5.662845087051392,
      "epoch": 2.292861311028983,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004468915488695777,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2136337175965309,
      "num_tokens": 51115329.0,
      "step": 29470
    },
    {
      "entropy": 5.611136531829834,
      "epoch": 2.2932503404007,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044687366378562036,
      "loss": 4.9591,
      "mean_token_accuracy": 0.20878891348838807,
      "num_tokens": 51124215.0,
      "step": 29475
    },
    {
      "entropy": 5.706327056884765,
      "epoch": 2.293639369772418,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044685577609375373,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2026585891842842,
      "num_tokens": 51132528.0,
      "step": 29480
    },
    {
      "entropy": 5.734607124328614,
      "epoch": 2.2940283991441355,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004468378857942492,
      "loss": 5.0698,
      "mean_token_accuracy": 0.1999109134078026,
      "num_tokens": 51140726.0,
      "step": 29485
    },
    {
      "entropy": 5.6961338996887205,
      "epoch": 2.294417428515853,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044681999288737826,
      "loss": 4.9684,
      "mean_token_accuracy": 0.20337729454040526,
      "num_tokens": 51149312.0,
      "step": 29490
    },
    {
      "entropy": 5.792264747619629,
      "epoch": 2.2948064578875704,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044680209737341244,
      "loss": 5.0803,
      "mean_token_accuracy": 0.1969618633389473,
      "num_tokens": 51157464.0,
      "step": 29495
    },
    {
      "entropy": 5.631439924240112,
      "epoch": 2.295195487259288,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004467841992526233,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2160213366150856,
      "num_tokens": 51166331.0,
      "step": 29500
    },
    {
      "entropy": 5.738448762893677,
      "epoch": 2.295584516631006,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044676629852528226,
      "loss": 5.0029,
      "mean_token_accuracy": 0.20881515145301818,
      "num_tokens": 51174404.0,
      "step": 29505
    },
    {
      "entropy": 5.649108982086181,
      "epoch": 2.295973546002723,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044674839519166104,
      "loss": 4.8326,
      "mean_token_accuracy": 0.21302674412727357,
      "num_tokens": 51182650.0,
      "step": 29510
    },
    {
      "entropy": 5.7233813285827635,
      "epoch": 2.2963625753744408,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044673048925203136,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20434564501047134,
      "num_tokens": 51192107.0,
      "step": 29515
    },
    {
      "entropy": 5.65120301246643,
      "epoch": 2.2967516047461585,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044671258070666476,
      "loss": 4.9612,
      "mean_token_accuracy": 0.20321808606386185,
      "num_tokens": 51200506.0,
      "step": 29520
    },
    {
      "entropy": 5.681089305877686,
      "epoch": 2.2971406341178757,
      "grad_norm": 1.125,
      "learning_rate": 0.0004466946695558331,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21102296859025954,
      "num_tokens": 51209442.0,
      "step": 29525
    },
    {
      "entropy": 5.698072004318237,
      "epoch": 2.2975296634895934,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044667675579980817,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2095807358622551,
      "num_tokens": 51217520.0,
      "step": 29530
    },
    {
      "entropy": 5.7934681415557865,
      "epoch": 2.297918692861311,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004466588394388616,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19809847474098205,
      "num_tokens": 51226228.0,
      "step": 29535
    },
    {
      "entropy": 5.787674522399902,
      "epoch": 2.2983077222330284,
      "grad_norm": 1.125,
      "learning_rate": 0.0004466409204732654,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20415193289518357,
      "num_tokens": 51235271.0,
      "step": 29540
    },
    {
      "entropy": 5.712382698059082,
      "epoch": 2.298696751604746,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044662299890329144,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20654861330986024,
      "num_tokens": 51243927.0,
      "step": 29545
    },
    {
      "entropy": 5.736461877822876,
      "epoch": 2.2990857809764638,
      "grad_norm": 1.5,
      "learning_rate": 0.0004466050747292116,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20454791486263274,
      "num_tokens": 51253097.0,
      "step": 29550
    },
    {
      "entropy": 5.758703184127808,
      "epoch": 2.2994748103481815,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000446587147951298,
      "loss": 4.9935,
      "mean_token_accuracy": 0.20700860768556595,
      "num_tokens": 51261275.0,
      "step": 29555
    },
    {
      "entropy": 5.771856260299683,
      "epoch": 2.2998638397198987,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004465692185698224,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20103539675474166,
      "num_tokens": 51269698.0,
      "step": 29560
    },
    {
      "entropy": 5.698578929901123,
      "epoch": 2.3002528690916164,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044655128658505717,
      "loss": 4.9224,
      "mean_token_accuracy": 0.21285854130983353,
      "num_tokens": 51277850.0,
      "step": 29565
    },
    {
      "entropy": 5.705301523208618,
      "epoch": 2.300641898463334,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004465333519972741,
      "loss": 4.939,
      "mean_token_accuracy": 0.20702046006917954,
      "num_tokens": 51285919.0,
      "step": 29570
    },
    {
      "entropy": 5.689915990829467,
      "epoch": 2.3010309278350514,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004465154148067454,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20669616460800172,
      "num_tokens": 51294403.0,
      "step": 29575
    },
    {
      "entropy": 5.801476287841797,
      "epoch": 2.301419957206769,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044649747501374336,
      "loss": 5.0426,
      "mean_token_accuracy": 0.19846068620681762,
      "num_tokens": 51302820.0,
      "step": 29580
    },
    {
      "entropy": 5.701091146469116,
      "epoch": 2.3018089865784868,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004464795326185401,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20767954140901565,
      "num_tokens": 51311385.0,
      "step": 29585
    },
    {
      "entropy": 5.756893873214722,
      "epoch": 2.3021980159502045,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004464615876214078,
      "loss": 5.0088,
      "mean_token_accuracy": 0.20564890056848525,
      "num_tokens": 51320928.0,
      "step": 29590
    },
    {
      "entropy": 5.663777256011963,
      "epoch": 2.3025870453219217,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004464436400226188,
      "loss": 4.9253,
      "mean_token_accuracy": 0.20751966685056686,
      "num_tokens": 51329384.0,
      "step": 29595
    },
    {
      "entropy": 5.684647274017334,
      "epoch": 2.3029760746936394,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004464256898224455,
      "loss": 4.9958,
      "mean_token_accuracy": 0.19487207084894181,
      "num_tokens": 51337765.0,
      "step": 29600
    },
    {
      "entropy": 5.740240001678467,
      "epoch": 2.303365104065357,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004464077370211602,
      "loss": 4.9793,
      "mean_token_accuracy": 0.20427297204732894,
      "num_tokens": 51345947.0,
      "step": 29605
    },
    {
      "entropy": 5.659935092926025,
      "epoch": 2.3037541334370744,
      "grad_norm": 1.25,
      "learning_rate": 0.00044638978161903533,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2043631136417389,
      "num_tokens": 51354742.0,
      "step": 29610
    },
    {
      "entropy": 5.633141374588012,
      "epoch": 2.304143162808792,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004463718236163433,
      "loss": 4.928,
      "mean_token_accuracy": 0.2093992128968239,
      "num_tokens": 51364208.0,
      "step": 29615
    },
    {
      "entropy": 5.741364812850952,
      "epoch": 2.3045321921805098,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044635386301335666,
      "loss": 5.0875,
      "mean_token_accuracy": 0.19642352908849717,
      "num_tokens": 51373081.0,
      "step": 29620
    },
    {
      "entropy": 5.6877683162689205,
      "epoch": 2.304921221552227,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004463358998103478,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20852137058973313,
      "num_tokens": 51382561.0,
      "step": 29625
    },
    {
      "entropy": 5.7853258609771725,
      "epoch": 2.3053102509239447,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004463179340075894,
      "loss": 5.055,
      "mean_token_accuracy": 0.20002379268407822,
      "num_tokens": 51391671.0,
      "step": 29630
    },
    {
      "entropy": 5.7338306427001955,
      "epoch": 2.3056992802956624,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004462999656053541,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21088672280311585,
      "num_tokens": 51400727.0,
      "step": 29635
    },
    {
      "entropy": 5.740619659423828,
      "epoch": 2.3060883096673797,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044628199460391445,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20177623331546785,
      "num_tokens": 51409943.0,
      "step": 29640
    },
    {
      "entropy": 5.729908847808838,
      "epoch": 2.3064773390390974,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044626402100354307,
      "loss": 5.0985,
      "mean_token_accuracy": 0.19475341141223906,
      "num_tokens": 51418511.0,
      "step": 29645
    },
    {
      "entropy": 5.6736303806304935,
      "epoch": 2.306866368410815,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004462460448045129,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2131457194685936,
      "num_tokens": 51426304.0,
      "step": 29650
    },
    {
      "entropy": 5.760821676254272,
      "epoch": 2.3072553977825327,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044622806600709645,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20432318449020387,
      "num_tokens": 51434779.0,
      "step": 29655
    },
    {
      "entropy": 5.724936771392822,
      "epoch": 2.30764442715425,
      "grad_norm": 1.125,
      "learning_rate": 0.0004462100846115667,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20824972093105315,
      "num_tokens": 51444275.0,
      "step": 29660
    },
    {
      "entropy": 5.727829551696777,
      "epoch": 2.3080334565259677,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004461921006181964,
      "loss": 5.0311,
      "mean_token_accuracy": 0.1954103097319603,
      "num_tokens": 51453149.0,
      "step": 29665
    },
    {
      "entropy": 5.685744762420654,
      "epoch": 2.3084224858976854,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004461741140272584,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21351730823516846,
      "num_tokens": 51462302.0,
      "step": 29670
    },
    {
      "entropy": 5.792515563964844,
      "epoch": 2.3088115152694026,
      "grad_norm": 1.25,
      "learning_rate": 0.0004461561248390257,
      "loss": 5.0678,
      "mean_token_accuracy": 0.19730890095233916,
      "num_tokens": 51471186.0,
      "step": 29675
    },
    {
      "entropy": 5.770266914367676,
      "epoch": 2.3092005446411203,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004461381330537713,
      "loss": 5.0126,
      "mean_token_accuracy": 0.20588287264108657,
      "num_tokens": 51479868.0,
      "step": 29680
    },
    {
      "entropy": 5.661131191253662,
      "epoch": 2.309589574012838,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000446120138671768,
      "loss": 4.9074,
      "mean_token_accuracy": 0.21280314326286315,
      "num_tokens": 51489118.0,
      "step": 29685
    },
    {
      "entropy": 5.665546560287476,
      "epoch": 2.3099786033845557,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044610214169328913,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20723063349723816,
      "num_tokens": 51497933.0,
      "step": 29690
    },
    {
      "entropy": 5.765706396102905,
      "epoch": 2.310367632756273,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004460841421186075,
      "loss": 5.0161,
      "mean_token_accuracy": 0.20125896334648133,
      "num_tokens": 51507046.0,
      "step": 29695
    },
    {
      "entropy": 5.764743757247925,
      "epoch": 2.3107566621279907,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004460661399479963,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20398916602134703,
      "num_tokens": 51516558.0,
      "step": 29700
    },
    {
      "entropy": 5.665904331207275,
      "epoch": 2.3111456914997084,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044604813518172876,
      "loss": 4.8971,
      "mean_token_accuracy": 0.21512556225061416,
      "num_tokens": 51524411.0,
      "step": 29705
    },
    {
      "entropy": 5.70937147140503,
      "epoch": 2.3115347208714256,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044603012782007796,
      "loss": 5.0054,
      "mean_token_accuracy": 0.19421593993902206,
      "num_tokens": 51533121.0,
      "step": 29710
    },
    {
      "entropy": 5.738081550598144,
      "epoch": 2.3119237502431433,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004460121178633172,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20278743356466294,
      "num_tokens": 51541417.0,
      "step": 29715
    },
    {
      "entropy": 5.763906764984131,
      "epoch": 2.312312779614861,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044599410531171986,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2149574115872383,
      "num_tokens": 51549444.0,
      "step": 29720
    },
    {
      "entropy": 5.720767974853516,
      "epoch": 2.3127018089865787,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000445976090165559,
      "loss": 5.0728,
      "mean_token_accuracy": 0.2034244030714035,
      "num_tokens": 51558090.0,
      "step": 29725
    },
    {
      "entropy": 5.774804592132568,
      "epoch": 2.313090838358296,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004459580724251082,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20004751831293105,
      "num_tokens": 51566794.0,
      "step": 29730
    },
    {
      "entropy": 5.674492073059082,
      "epoch": 2.3134798677300137,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044594005209064064,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21250143498182297,
      "num_tokens": 51575411.0,
      "step": 29735
    },
    {
      "entropy": 5.721498775482178,
      "epoch": 2.313868897101731,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044592202916243003,
      "loss": 5.023,
      "mean_token_accuracy": 0.20184371769428253,
      "num_tokens": 51584438.0,
      "step": 29740
    },
    {
      "entropy": 5.723281812667847,
      "epoch": 2.3142579264734486,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004459040036407495,
      "loss": 4.9997,
      "mean_token_accuracy": 0.205299673974514,
      "num_tokens": 51593517.0,
      "step": 29745
    },
    {
      "entropy": 5.752590751647949,
      "epoch": 2.3146469558451663,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004458859755258729,
      "loss": 5.0342,
      "mean_token_accuracy": 0.1999933212995529,
      "num_tokens": 51602312.0,
      "step": 29750
    },
    {
      "entropy": 5.680380868911743,
      "epoch": 2.315035985216884,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044586794481807357,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2157074123620987,
      "num_tokens": 51611021.0,
      "step": 29755
    },
    {
      "entropy": 5.672012138366699,
      "epoch": 2.3154250145886013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044584991151762506,
      "loss": 5.0317,
      "mean_token_accuracy": 0.19843609780073165,
      "num_tokens": 51620322.0,
      "step": 29760
    },
    {
      "entropy": 5.716323757171631,
      "epoch": 2.315814043960319,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004458318756248011,
      "loss": 4.961,
      "mean_token_accuracy": 0.2111661598086357,
      "num_tokens": 51628626.0,
      "step": 29765
    },
    {
      "entropy": 5.693981266021728,
      "epoch": 2.3162030733320367,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044581383713987547,
      "loss": 4.9508,
      "mean_token_accuracy": 0.20812192261219026,
      "num_tokens": 51637868.0,
      "step": 29770
    },
    {
      "entropy": 5.735709762573242,
      "epoch": 2.316592102703754,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004457957960631216,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2017674505710602,
      "num_tokens": 51645756.0,
      "step": 29775
    },
    {
      "entropy": 5.716886758804321,
      "epoch": 2.3169811320754716,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004457777523948134,
      "loss": 5.05,
      "mean_token_accuracy": 0.19895590394735335,
      "num_tokens": 51654959.0,
      "step": 29780
    },
    {
      "entropy": 5.7394468784332275,
      "epoch": 2.3173701614471893,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004457597061352247,
      "loss": 5.0285,
      "mean_token_accuracy": 0.20409662276506424,
      "num_tokens": 51663208.0,
      "step": 29785
    },
    {
      "entropy": 5.703838443756103,
      "epoch": 2.317759190818907,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004457416572846291,
      "loss": 5.0421,
      "mean_token_accuracy": 0.19635406285524368,
      "num_tokens": 51671627.0,
      "step": 29790
    },
    {
      "entropy": 5.61381893157959,
      "epoch": 2.3181482201906243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004457236058433008,
      "loss": 4.9231,
      "mean_token_accuracy": 0.20901954919099808,
      "num_tokens": 51680870.0,
      "step": 29795
    },
    {
      "entropy": 5.732099628448486,
      "epoch": 2.318537249562342,
      "grad_norm": 1.25,
      "learning_rate": 0.00044570555181151333,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19941259920597076,
      "num_tokens": 51690566.0,
      "step": 29800
    },
    {
      "entropy": 5.773785734176636,
      "epoch": 2.3189262789340597,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000445687495189541,
      "loss": 5.0617,
      "mean_token_accuracy": 0.20187411159276963,
      "num_tokens": 51698792.0,
      "step": 29805
    },
    {
      "entropy": 5.74344220161438,
      "epoch": 2.319315308305777,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044566943597765745,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20119127333164216,
      "num_tokens": 51707821.0,
      "step": 29810
    },
    {
      "entropy": 5.660280752182007,
      "epoch": 2.3197043376774946,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044565137417613694,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21025829315185546,
      "num_tokens": 51716802.0,
      "step": 29815
    },
    {
      "entropy": 5.691473436355591,
      "epoch": 2.3200933670492123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004456333097852534,
      "loss": 4.987,
      "mean_token_accuracy": 0.2051575616002083,
      "num_tokens": 51725644.0,
      "step": 29820
    },
    {
      "entropy": 5.8301825523376465,
      "epoch": 2.32048239642093,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044561524280528104,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19584852904081346,
      "num_tokens": 51734072.0,
      "step": 29825
    },
    {
      "entropy": 5.788702583312988,
      "epoch": 2.3208714257926473,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044559717323649383,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2095063343644142,
      "num_tokens": 51743112.0,
      "step": 29830
    },
    {
      "entropy": 5.726411390304565,
      "epoch": 2.321260455164365,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004455791010791662,
      "loss": 5.0186,
      "mean_token_accuracy": 0.19476889818906784,
      "num_tokens": 51752535.0,
      "step": 29835
    },
    {
      "entropy": 5.690359210968017,
      "epoch": 2.3216494845360827,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004455610263335721,
      "loss": 4.977,
      "mean_token_accuracy": 0.20556507259607315,
      "num_tokens": 51760595.0,
      "step": 29840
    },
    {
      "entropy": 5.747605085372925,
      "epoch": 2.3220385139078,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004455429489999859,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20572015047073364,
      "num_tokens": 51769156.0,
      "step": 29845
    },
    {
      "entropy": 5.731538486480713,
      "epoch": 2.3224275432795176,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044552486907868194,
      "loss": 4.9753,
      "mean_token_accuracy": 0.20061874985694886,
      "num_tokens": 51778091.0,
      "step": 29850
    },
    {
      "entropy": 5.757988548278808,
      "epoch": 2.3228165726512353,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044550678656993454,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20715163499116898,
      "num_tokens": 51787207.0,
      "step": 29855
    },
    {
      "entropy": 5.6261577129364015,
      "epoch": 2.3232056020229526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000445488701474018,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2144611269235611,
      "num_tokens": 51796158.0,
      "step": 29860
    },
    {
      "entropy": 5.6302642822265625,
      "epoch": 2.3235946313946703,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004454706137912067,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2176444038748741,
      "num_tokens": 51803762.0,
      "step": 29865
    },
    {
      "entropy": 5.742434453964234,
      "epoch": 2.323983660766388,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004454525235217753,
      "loss": 5.024,
      "mean_token_accuracy": 0.20313291847705842,
      "num_tokens": 51812610.0,
      "step": 29870
    },
    {
      "entropy": 5.6924785614013675,
      "epoch": 2.324372690138105,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044543443066599807,
      "loss": 5.0067,
      "mean_token_accuracy": 0.20187889486551286,
      "num_tokens": 51821398.0,
      "step": 29875
    },
    {
      "entropy": 5.703952503204346,
      "epoch": 2.324761719509823,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004454163352241498,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2026186004281044,
      "num_tokens": 51829945.0,
      "step": 29880
    },
    {
      "entropy": 5.710974311828613,
      "epoch": 2.3251507488815406,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044539823719650463,
      "loss": 4.986,
      "mean_token_accuracy": 0.20879193097352983,
      "num_tokens": 51838816.0,
      "step": 29885
    },
    {
      "entropy": 5.685320568084717,
      "epoch": 2.3255397782532583,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004453801365833376,
      "loss": 5.0154,
      "mean_token_accuracy": 0.20976658761501313,
      "num_tokens": 51847775.0,
      "step": 29890
    },
    {
      "entropy": 5.744823169708252,
      "epoch": 2.3259288076249756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044536203338492317,
      "loss": 4.9876,
      "mean_token_accuracy": 0.2036099076271057,
      "num_tokens": 51856934.0,
      "step": 29895
    },
    {
      "entropy": 5.783577728271484,
      "epoch": 2.3263178369966933,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044534392760153596,
      "loss": 5.0434,
      "mean_token_accuracy": 0.1947759598493576,
      "num_tokens": 51865393.0,
      "step": 29900
    },
    {
      "entropy": 5.688048315048218,
      "epoch": 2.326706866368411,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004453258192334508,
      "loss": 4.9546,
      "mean_token_accuracy": 0.2104189857840538,
      "num_tokens": 51873783.0,
      "step": 29905
    },
    {
      "entropy": 5.654390954971314,
      "epoch": 2.327095895740128,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004453077082809425,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19473850280046462,
      "num_tokens": 51882934.0,
      "step": 29910
    },
    {
      "entropy": 5.7023172855377195,
      "epoch": 2.327484925111846,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044528959474428575,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20492217987775802,
      "num_tokens": 51891217.0,
      "step": 29915
    },
    {
      "entropy": 5.805389451980591,
      "epoch": 2.3278739544835636,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004452714786237555,
      "loss": 5.054,
      "mean_token_accuracy": 0.20165701508522033,
      "num_tokens": 51900550.0,
      "step": 29920
    },
    {
      "entropy": 5.7403521060943605,
      "epoch": 2.3282629838552813,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004452533599196265,
      "loss": 5.015,
      "mean_token_accuracy": 0.1995408356189728,
      "num_tokens": 51908436.0,
      "step": 29925
    },
    {
      "entropy": 5.682050085067749,
      "epoch": 2.3286520132269986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044523523863217374,
      "loss": 5.032,
      "mean_token_accuracy": 0.2118928089737892,
      "num_tokens": 51917687.0,
      "step": 29930
    },
    {
      "entropy": 5.705183219909668,
      "epoch": 2.3290410425987162,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004452171147616723,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20223670750856398,
      "num_tokens": 51927011.0,
      "step": 29935
    },
    {
      "entropy": 5.79772720336914,
      "epoch": 2.329430071970434,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000445198988308397,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19700517803430556,
      "num_tokens": 51936255.0,
      "step": 29940
    },
    {
      "entropy": 5.722238063812256,
      "epoch": 2.329819101342151,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044518085927262293,
      "loss": 5.0299,
      "mean_token_accuracy": 0.19975855946540833,
      "num_tokens": 51944919.0,
      "step": 29945
    },
    {
      "entropy": 5.738206481933593,
      "epoch": 2.330208130713869,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004451627276546252,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20174067616462707,
      "num_tokens": 51952963.0,
      "step": 29950
    },
    {
      "entropy": 5.716019248962402,
      "epoch": 2.3305971600855866,
      "grad_norm": 1.1875,
      "learning_rate": 0.000445144593454679,
      "loss": 4.9416,
      "mean_token_accuracy": 0.21096449345350266,
      "num_tokens": 51961493.0,
      "step": 29955
    },
    {
      "entropy": 5.674034833908081,
      "epoch": 2.330986189457304,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004451264566730593,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2033722683787346,
      "num_tokens": 51969911.0,
      "step": 29960
    },
    {
      "entropy": 5.6917815685272215,
      "epoch": 2.3313752188290215,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044510831731004146,
      "loss": 4.956,
      "mean_token_accuracy": 0.20430531203746796,
      "num_tokens": 51978936.0,
      "step": 29965
    },
    {
      "entropy": 5.7915942668914795,
      "epoch": 2.3317642482007392,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004450901753659007,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2002222254872322,
      "num_tokens": 51987373.0,
      "step": 29970
    },
    {
      "entropy": 5.7442262172698975,
      "epoch": 2.3321532775724565,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044507203084091215,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20064566731452943,
      "num_tokens": 51995948.0,
      "step": 29975
    },
    {
      "entropy": 5.716240644454956,
      "epoch": 2.332542306944174,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004450538837353513,
      "loss": 5.0335,
      "mean_token_accuracy": 0.19856160134077072,
      "num_tokens": 52004623.0,
      "step": 29980
    },
    {
      "entropy": 5.734252882003784,
      "epoch": 2.332931336315892,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044503573404949343,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20958548039197922,
      "num_tokens": 52012971.0,
      "step": 29985
    },
    {
      "entropy": 5.761873006820679,
      "epoch": 2.3333203656876096,
      "grad_norm": 1.25,
      "learning_rate": 0.0004450175817836139,
      "loss": 5.0643,
      "mean_token_accuracy": 0.19934442043304443,
      "num_tokens": 52021462.0,
      "step": 29990
    },
    {
      "entropy": 5.689078426361084,
      "epoch": 2.333709395059327,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044499942693798823,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20151477009058,
      "num_tokens": 52030494.0,
      "step": 29995
    },
    {
      "entropy": 5.6934003829956055,
      "epoch": 2.3340984244310445,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004449812695128918,
      "loss": 4.941,
      "mean_token_accuracy": 0.2089556038379669,
      "num_tokens": 52039120.0,
      "step": 30000
    },
    {
      "epoch": 2.3340984244310445,
      "eval_entropy": 5.46833552350438,
      "eval_loss": 5.083688735961914,
      "eval_mean_token_accuracy": 0.21000343827213427,
      "eval_num_tokens": 52039120.0,
      "eval_runtime": 21.796,
      "eval_samples_per_second": 1906.677,
      "eval_steps_per_second": 238.346,
      "step": 30000
    },
    {
      "entropy": 5.796833562850952,
      "epoch": 2.3344874538027622,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044496310950860015,
      "loss": 5.0954,
      "mean_token_accuracy": 0.19699225574731827,
      "num_tokens": 52047248.0,
      "step": 30005
    },
    {
      "entropy": 5.721827507019043,
      "epoch": 2.3348764831744795,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044494494692538886,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2074394181370735,
      "num_tokens": 52056359.0,
      "step": 30010
    },
    {
      "entropy": 5.758582401275635,
      "epoch": 2.335265512546197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044492678176353347,
      "loss": 5.0123,
      "mean_token_accuracy": 0.20021031945943832,
      "num_tokens": 52065042.0,
      "step": 30015
    },
    {
      "entropy": 5.653558969497681,
      "epoch": 2.335654541917915,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044490861402330967,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2180691808462143,
      "num_tokens": 52073830.0,
      "step": 30020
    },
    {
      "entropy": 5.601017713546753,
      "epoch": 2.3360435712896326,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004448904437049931,
      "loss": 4.8691,
      "mean_token_accuracy": 0.21003260612487792,
      "num_tokens": 52082605.0,
      "step": 30025
    },
    {
      "entropy": 5.718303680419922,
      "epoch": 2.33643260066135,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004448722708088594,
      "loss": 5.1415,
      "mean_token_accuracy": 0.19393166154623032,
      "num_tokens": 52091997.0,
      "step": 30030
    },
    {
      "entropy": 5.735299730300904,
      "epoch": 2.3368216300330675,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004448540953351844,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2058476150035858,
      "num_tokens": 52100690.0,
      "step": 30035
    },
    {
      "entropy": 5.743722867965698,
      "epoch": 2.3372106594047852,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004448359172842439,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20715974718332292,
      "num_tokens": 52109310.0,
      "step": 30040
    },
    {
      "entropy": 5.762193489074707,
      "epoch": 2.3375996887765025,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044481773665631355,
      "loss": 5.0174,
      "mean_token_accuracy": 0.2137180119752884,
      "num_tokens": 52118662.0,
      "step": 30045
    },
    {
      "entropy": 5.658874559402466,
      "epoch": 2.33798871814822,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004447995534516695,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21392421722412108,
      "num_tokens": 52127098.0,
      "step": 30050
    },
    {
      "entropy": 5.686005973815918,
      "epoch": 2.338377747519938,
      "grad_norm": 1.25,
      "learning_rate": 0.00044478136767058733,
      "loss": 5.0323,
      "mean_token_accuracy": 0.19525890201330184,
      "num_tokens": 52135539.0,
      "step": 30055
    },
    {
      "entropy": 5.775244426727295,
      "epoch": 2.338766776891655,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004447631793133432,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20425243824720382,
      "num_tokens": 52144169.0,
      "step": 30060
    },
    {
      "entropy": 5.7356476306915285,
      "epoch": 2.339155806263373,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004447449883802131,
      "loss": 5.039,
      "mean_token_accuracy": 0.19800032526254654,
      "num_tokens": 52152679.0,
      "step": 30065
    },
    {
      "entropy": 5.6290160655975345,
      "epoch": 2.3395448356350905,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044472679487147295,
      "loss": 4.973,
      "mean_token_accuracy": 0.2056146740913391,
      "num_tokens": 52161407.0,
      "step": 30070
    },
    {
      "entropy": 5.8137890815734865,
      "epoch": 2.339933865006808,
      "grad_norm": 1.125,
      "learning_rate": 0.00044470859878739877,
      "loss": 5.0518,
      "mean_token_accuracy": 0.19473549574613572,
      "num_tokens": 52170285.0,
      "step": 30075
    },
    {
      "entropy": 5.751782703399658,
      "epoch": 2.3403228943785255,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044469040012826676,
      "loss": 4.8947,
      "mean_token_accuracy": 0.21114516258239746,
      "num_tokens": 52178813.0,
      "step": 30080
    },
    {
      "entropy": 5.676619672775269,
      "epoch": 2.340711923750243,
      "grad_norm": 1.25,
      "learning_rate": 0.00044467219889435304,
      "loss": 4.9048,
      "mean_token_accuracy": 0.21357868760824203,
      "num_tokens": 52187205.0,
      "step": 30085
    },
    {
      "entropy": 5.692780351638794,
      "epoch": 2.341100953121961,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004446539950859337,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20415353626012803,
      "num_tokens": 52195318.0,
      "step": 30090
    },
    {
      "entropy": 5.698045969009399,
      "epoch": 2.341489982493678,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044463578870328506,
      "loss": 4.976,
      "mean_token_accuracy": 0.2029847741127014,
      "num_tokens": 52204093.0,
      "step": 30095
    },
    {
      "entropy": 5.686465930938721,
      "epoch": 2.341879011865396,
      "grad_norm": 1.25,
      "learning_rate": 0.0004446175797466834,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20910148024559022,
      "num_tokens": 52212319.0,
      "step": 30100
    },
    {
      "entropy": 5.669716262817383,
      "epoch": 2.3422680412371135,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044459936821640485,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20232271999120713,
      "num_tokens": 52221878.0,
      "step": 30105
    },
    {
      "entropy": 5.656469297409058,
      "epoch": 2.3426570706088308,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004445811541127258,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21309261173009872,
      "num_tokens": 52230952.0,
      "step": 30110
    },
    {
      "entropy": 5.707837963104248,
      "epoch": 2.3430460999805485,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044456293743592274,
      "loss": 5.0242,
      "mean_token_accuracy": 0.2099805146455765,
      "num_tokens": 52239461.0,
      "step": 30115
    },
    {
      "entropy": 5.708162832260132,
      "epoch": 2.343435129352266,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044454471818627196,
      "loss": 4.9855,
      "mean_token_accuracy": 0.20382254272699357,
      "num_tokens": 52248478.0,
      "step": 30120
    },
    {
      "entropy": 5.729603099822998,
      "epoch": 2.343824158723984,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044452649636404995,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2047068417072296,
      "num_tokens": 52256901.0,
      "step": 30125
    },
    {
      "entropy": 5.731555795669555,
      "epoch": 2.344213188095701,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004445082719695332,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2082977309823036,
      "num_tokens": 52265249.0,
      "step": 30130
    },
    {
      "entropy": 5.757996940612793,
      "epoch": 2.344602217467419,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044449004500299815,
      "loss": 5.0327,
      "mean_token_accuracy": 0.200745390355587,
      "num_tokens": 52274777.0,
      "step": 30135
    },
    {
      "entropy": 5.741820192337036,
      "epoch": 2.3449912468391365,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044447181546472146,
      "loss": 5.0467,
      "mean_token_accuracy": 0.20020510852336884,
      "num_tokens": 52283392.0,
      "step": 30140
    },
    {
      "entropy": 5.63641242980957,
      "epoch": 2.3453802762108538,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044445358335497975,
      "loss": 4.9525,
      "mean_token_accuracy": 0.20334971845149993,
      "num_tokens": 52291875.0,
      "step": 30145
    },
    {
      "entropy": 5.719304323196411,
      "epoch": 2.3457693055825715,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004444353486740496,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20716226398944854,
      "num_tokens": 52300194.0,
      "step": 30150
    },
    {
      "entropy": 5.755153322219849,
      "epoch": 2.346158334954289,
      "grad_norm": 1.25,
      "learning_rate": 0.0004444171114222078,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20190504789352418,
      "num_tokens": 52309147.0,
      "step": 30155
    },
    {
      "entropy": 5.788867855072022,
      "epoch": 2.346547364326007,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044439887159973096,
      "loss": 5.1041,
      "mean_token_accuracy": 0.20365861654281617,
      "num_tokens": 52317502.0,
      "step": 30160
    },
    {
      "entropy": 5.691752529144287,
      "epoch": 2.346936393697724,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004443806292068958,
      "loss": 4.8973,
      "mean_token_accuracy": 0.21190401464700698,
      "num_tokens": 52325539.0,
      "step": 30165
    },
    {
      "entropy": 5.747013807296753,
      "epoch": 2.347325423069442,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004443623842439792,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20385870039463044,
      "num_tokens": 52333676.0,
      "step": 30170
    },
    {
      "entropy": 5.716862297058105,
      "epoch": 2.347714452441159,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000444344136711258,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21453170478343964,
      "num_tokens": 52342706.0,
      "step": 30175
    },
    {
      "entropy": 5.687949800491333,
      "epoch": 2.3481034818128768,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044432588660900905,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20466179996728898,
      "num_tokens": 52351409.0,
      "step": 30180
    },
    {
      "entropy": 5.69847559928894,
      "epoch": 2.3484925111845945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004443076339375093,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2092856779694557,
      "num_tokens": 52359781.0,
      "step": 30185
    },
    {
      "entropy": 5.736768674850464,
      "epoch": 2.348881540556312,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044428937869703575,
      "loss": 4.9756,
      "mean_token_accuracy": 0.20476006865501403,
      "num_tokens": 52368236.0,
      "step": 30190
    },
    {
      "entropy": 5.7104979991912845,
      "epoch": 2.3492705699280294,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044427112088786525,
      "loss": 5.044,
      "mean_token_accuracy": 0.2039959415793419,
      "num_tokens": 52376449.0,
      "step": 30195
    },
    {
      "entropy": 5.688129234313965,
      "epoch": 2.349659599299747,
      "grad_norm": 1.25,
      "learning_rate": 0.0004442528605102749,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2053952008485794,
      "num_tokens": 52384844.0,
      "step": 30200
    },
    {
      "entropy": 5.77655668258667,
      "epoch": 2.350048628671465,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004442345975645418,
      "loss": 5.1077,
      "mean_token_accuracy": 0.20078441351652146,
      "num_tokens": 52393810.0,
      "step": 30205
    },
    {
      "entropy": 5.745203638076783,
      "epoch": 2.350437658043182,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004442163320509431,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20884886234998704,
      "num_tokens": 52402081.0,
      "step": 30210
    },
    {
      "entropy": 5.652178144454956,
      "epoch": 2.3508266874148998,
      "grad_norm": 1.5,
      "learning_rate": 0.0004441980639697558,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20009441077709197,
      "num_tokens": 52411671.0,
      "step": 30215
    },
    {
      "entropy": 5.6855055809021,
      "epoch": 2.3512157167866174,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004441797933212573,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20755104571580887,
      "num_tokens": 52420255.0,
      "step": 30220
    },
    {
      "entropy": 5.797997760772705,
      "epoch": 2.351604746158335,
      "grad_norm": 1.125,
      "learning_rate": 0.0004441615201057247,
      "loss": 5.1556,
      "mean_token_accuracy": 0.18963198065757753,
      "num_tokens": 52429682.0,
      "step": 30225
    },
    {
      "entropy": 5.758996772766113,
      "epoch": 2.3519937755300524,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004441432443234352,
      "loss": 4.9813,
      "mean_token_accuracy": 0.20145370364189147,
      "num_tokens": 52438719.0,
      "step": 30230
    },
    {
      "entropy": 5.650208520889282,
      "epoch": 2.35238280490177,
      "grad_norm": 1.125,
      "learning_rate": 0.00044412496597466626,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2106275036931038,
      "num_tokens": 52447543.0,
      "step": 30235
    },
    {
      "entropy": 5.688118267059326,
      "epoch": 2.352771834273488,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004441066850596951,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20508147776126862,
      "num_tokens": 52456606.0,
      "step": 30240
    },
    {
      "entropy": 5.74537410736084,
      "epoch": 2.353160863645205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004440884015787992,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20507865250110627,
      "num_tokens": 52465222.0,
      "step": 30245
    },
    {
      "entropy": 5.71884765625,
      "epoch": 2.3535498930169227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004440701155322558,
      "loss": 5.039,
      "mean_token_accuracy": 0.20404176265001298,
      "num_tokens": 52473955.0,
      "step": 30250
    },
    {
      "entropy": 5.694201898574829,
      "epoch": 2.3539389223886404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004440518269203427,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20405290722846986,
      "num_tokens": 52483786.0,
      "step": 30255
    },
    {
      "entropy": 5.801051425933838,
      "epoch": 2.354327951760358,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044403353574333715,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20028405636548996,
      "num_tokens": 52492292.0,
      "step": 30260
    },
    {
      "entropy": 5.715303564071656,
      "epoch": 2.3547169811320754,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044401524200151667,
      "loss": 5.0306,
      "mean_token_accuracy": 0.19436792135238648,
      "num_tokens": 52501267.0,
      "step": 30265
    },
    {
      "entropy": 5.790611457824707,
      "epoch": 2.355106010503793,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044399694569515895,
      "loss": 5.0591,
      "mean_token_accuracy": 0.2055249035358429,
      "num_tokens": 52509515.0,
      "step": 30270
    },
    {
      "entropy": 5.736593914031983,
      "epoch": 2.355495039875511,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044397864682454156,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20420722365379335,
      "num_tokens": 52517757.0,
      "step": 30275
    },
    {
      "entropy": 5.733224964141845,
      "epoch": 2.355884069247228,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004439603453899421,
      "loss": 5.0742,
      "mean_token_accuracy": 0.2039773643016815,
      "num_tokens": 52525996.0,
      "step": 30280
    },
    {
      "entropy": 5.656892251968384,
      "epoch": 2.3562730986189457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004439420413916384,
      "loss": 4.9438,
      "mean_token_accuracy": 0.20949291288852692,
      "num_tokens": 52534457.0,
      "step": 30285
    },
    {
      "entropy": 5.717184782028198,
      "epoch": 2.3566621279906634,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044392373482990817,
      "loss": 5.1037,
      "mean_token_accuracy": 0.19738477766513823,
      "num_tokens": 52543754.0,
      "step": 30290
    },
    {
      "entropy": 5.795910978317261,
      "epoch": 2.3570511573623807,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004439054257050291,
      "loss": 5.0275,
      "mean_token_accuracy": 0.20114080160856246,
      "num_tokens": 52552857.0,
      "step": 30295
    },
    {
      "entropy": 5.828827381134033,
      "epoch": 2.3574401867340984,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004438871140172789,
      "loss": 5.133,
      "mean_token_accuracy": 0.19353933334350587,
      "num_tokens": 52562760.0,
      "step": 30300
    },
    {
      "entropy": 5.69392409324646,
      "epoch": 2.357829216105816,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004438687997669357,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21247211694717408,
      "num_tokens": 52571111.0,
      "step": 30305
    },
    {
      "entropy": 5.655138826370239,
      "epoch": 2.3582182454775333,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044385048295427724,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21399367451667786,
      "num_tokens": 52579541.0,
      "step": 30310
    },
    {
      "entropy": 5.70272388458252,
      "epoch": 2.358607274849251,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004438321635795814,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20998730659484863,
      "num_tokens": 52588253.0,
      "step": 30315
    },
    {
      "entropy": 5.691138029098511,
      "epoch": 2.3589963042209687,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004438138416431262,
      "loss": 4.9806,
      "mean_token_accuracy": 0.19880460500717162,
      "num_tokens": 52597922.0,
      "step": 30320
    },
    {
      "entropy": 5.7377800941467285,
      "epoch": 2.3593853335926864,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044379551714518966,
      "loss": 4.9111,
      "mean_token_accuracy": 0.20879417806863784,
      "num_tokens": 52606167.0,
      "step": 30325
    },
    {
      "entropy": 5.726652908325195,
      "epoch": 2.3597743629644037,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044377719008604984,
      "loss": 5.1078,
      "mean_token_accuracy": 0.20196451544761657,
      "num_tokens": 52614534.0,
      "step": 30330
    },
    {
      "entropy": 5.739723205566406,
      "epoch": 2.3601633923361214,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004437588604659848,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2004109650850296,
      "num_tokens": 52623225.0,
      "step": 30335
    },
    {
      "entropy": 5.769060850143433,
      "epoch": 2.360552421707839,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004437405282852726,
      "loss": 5.0371,
      "mean_token_accuracy": 0.20414875000715255,
      "num_tokens": 52632025.0,
      "step": 30340
    },
    {
      "entropy": 5.685723781585693,
      "epoch": 2.3609414510795563,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044372219354419154,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22272738665342331,
      "num_tokens": 52640731.0,
      "step": 30345
    },
    {
      "entropy": 5.675121688842774,
      "epoch": 2.361330480451274,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004437038562430197,
      "loss": 4.9793,
      "mean_token_accuracy": 0.20451977252960205,
      "num_tokens": 52648831.0,
      "step": 30350
    },
    {
      "entropy": 5.75298810005188,
      "epoch": 2.3617195098229917,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044368551638203536,
      "loss": 5.0422,
      "mean_token_accuracy": 0.1990748390555382,
      "num_tokens": 52657699.0,
      "step": 30355
    },
    {
      "entropy": 5.7055260181427006,
      "epoch": 2.3621085391947094,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004436671739615168,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21120693534612656,
      "num_tokens": 52666418.0,
      "step": 30360
    },
    {
      "entropy": 5.680373001098633,
      "epoch": 2.3624975685664267,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044364882898174235,
      "loss": 5.009,
      "mean_token_accuracy": 0.2018944501876831,
      "num_tokens": 52675014.0,
      "step": 30365
    },
    {
      "entropy": 5.789453792572021,
      "epoch": 2.3628865979381444,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004436304814429903,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20092105865478516,
      "num_tokens": 52683691.0,
      "step": 30370
    },
    {
      "entropy": 5.77302565574646,
      "epoch": 2.363275627309862,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004436121313455391,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20131551921367646,
      "num_tokens": 52692317.0,
      "step": 30375
    },
    {
      "entropy": 5.688556814193726,
      "epoch": 2.3636646566815793,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004435937786896673,
      "loss": 5.0044,
      "mean_token_accuracy": 0.1977980524301529,
      "num_tokens": 52701523.0,
      "step": 30380
    },
    {
      "entropy": 5.749579191207886,
      "epoch": 2.364053686053297,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044357542347565323,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19970246106386186,
      "num_tokens": 52710436.0,
      "step": 30385
    },
    {
      "entropy": 5.769369649887085,
      "epoch": 2.3644427154250147,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004435570657037753,
      "loss": 5.0476,
      "mean_token_accuracy": 0.20139719694852828,
      "num_tokens": 52719186.0,
      "step": 30390
    },
    {
      "entropy": 5.731852006912232,
      "epoch": 2.364831744796732,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004435387053743123,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2043033555150032,
      "num_tokens": 52728185.0,
      "step": 30395
    },
    {
      "entropy": 5.735817193984985,
      "epoch": 2.3652207741684497,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044352034248754265,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20674290508031845,
      "num_tokens": 52737938.0,
      "step": 30400
    },
    {
      "entropy": 5.712808847427368,
      "epoch": 2.3656098035401674,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000443501977043745,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20584792047739028,
      "num_tokens": 52747005.0,
      "step": 30405
    },
    {
      "entropy": 5.639086818695068,
      "epoch": 2.3659988329118846,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004434836090431981,
      "loss": 4.9347,
      "mean_token_accuracy": 0.20374073684215546,
      "num_tokens": 52756251.0,
      "step": 30410
    },
    {
      "entropy": 5.698932456970215,
      "epoch": 2.3663878622836023,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004434652384861806,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20462170243263245,
      "num_tokens": 52764443.0,
      "step": 30415
    },
    {
      "entropy": 5.645833253860474,
      "epoch": 2.36677689165532,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004434468653729712,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2147284835577011,
      "num_tokens": 52772824.0,
      "step": 30420
    },
    {
      "entropy": 5.697416734695435,
      "epoch": 2.3671659210270377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044342848970384876,
      "loss": 5.0006,
      "mean_token_accuracy": 0.2041485548019409,
      "num_tokens": 52782743.0,
      "step": 30425
    },
    {
      "entropy": 5.801953744888306,
      "epoch": 2.367554950398755,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044341011147909206,
      "loss": 5.1315,
      "mean_token_accuracy": 0.19832820892333985,
      "num_tokens": 52791564.0,
      "step": 30430
    },
    {
      "entropy": 5.6965086460113525,
      "epoch": 2.3679439797704727,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044339173069897996,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21718985438346863,
      "num_tokens": 52800229.0,
      "step": 30435
    },
    {
      "entropy": 5.687645721435547,
      "epoch": 2.3683330091421904,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044337334736379143,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2090029940009117,
      "num_tokens": 52808342.0,
      "step": 30440
    },
    {
      "entropy": 5.697915840148926,
      "epoch": 2.3687220385139076,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004433549614738053,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20607690066099166,
      "num_tokens": 52817462.0,
      "step": 30445
    },
    {
      "entropy": 5.647302436828613,
      "epoch": 2.3691110678856253,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044333657302930056,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21117393672466278,
      "num_tokens": 52825567.0,
      "step": 30450
    },
    {
      "entropy": 5.71310248374939,
      "epoch": 2.369500097257343,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004433181820305564,
      "loss": 4.9659,
      "mean_token_accuracy": 0.20541804283857346,
      "num_tokens": 52835771.0,
      "step": 30455
    },
    {
      "entropy": 5.682458209991455,
      "epoch": 2.3698891266290607,
      "grad_norm": 1.25,
      "learning_rate": 0.00044329978847785156,
      "loss": 4.985,
      "mean_token_accuracy": 0.2053128093481064,
      "num_tokens": 52844344.0,
      "step": 30460
    },
    {
      "entropy": 5.740370559692383,
      "epoch": 2.370278156000778,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004432813923714654,
      "loss": 5.0155,
      "mean_token_accuracy": 0.20483268797397614,
      "num_tokens": 52852918.0,
      "step": 30465
    },
    {
      "entropy": 5.695824098587036,
      "epoch": 2.3706671853724957,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044326299371167695,
      "loss": 5.0377,
      "mean_token_accuracy": 0.205628003180027,
      "num_tokens": 52862142.0,
      "step": 30470
    },
    {
      "entropy": 5.728581190109253,
      "epoch": 2.3710562147442134,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044324459249876536,
      "loss": 5.0481,
      "mean_token_accuracy": 0.20105820745229722,
      "num_tokens": 52870595.0,
      "step": 30475
    },
    {
      "entropy": 5.762945604324341,
      "epoch": 2.3714452441159306,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004432261887330099,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20344586968421935,
      "num_tokens": 52879995.0,
      "step": 30480
    },
    {
      "entropy": 5.718740606307984,
      "epoch": 2.3718342734876483,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004432077824146898,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21418362408876418,
      "num_tokens": 52888341.0,
      "step": 30485
    },
    {
      "entropy": 5.639641141891479,
      "epoch": 2.372223302859366,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004431893735440843,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21180813908576965,
      "num_tokens": 52896617.0,
      "step": 30490
    },
    {
      "entropy": 5.649254894256591,
      "epoch": 2.3726123322310833,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004431709621214727,
      "loss": 4.9368,
      "mean_token_accuracy": 0.20986569672822952,
      "num_tokens": 52905071.0,
      "step": 30495
    },
    {
      "entropy": 5.6996125221252445,
      "epoch": 2.373001361602801,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044315254814713455,
      "loss": 4.921,
      "mean_token_accuracy": 0.21934307664632796,
      "num_tokens": 52913756.0,
      "step": 30500
    },
    {
      "entropy": 5.682266092300415,
      "epoch": 2.3733903909745186,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044313413162134894,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22312916815280914,
      "num_tokens": 52922895.0,
      "step": 30505
    },
    {
      "entropy": 5.724899482727051,
      "epoch": 2.373779420346236,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004431157125443958,
      "loss": 5.0461,
      "mean_token_accuracy": 0.2025214597582817,
      "num_tokens": 52931467.0,
      "step": 30510
    },
    {
      "entropy": 5.737158918380738,
      "epoch": 2.3741684497179536,
      "grad_norm": 1.171875,
      "learning_rate": 0.000443097290916554,
      "loss": 5.1032,
      "mean_token_accuracy": 0.19683476984500886,
      "num_tokens": 52941122.0,
      "step": 30515
    },
    {
      "entropy": 5.732145214080811,
      "epoch": 2.3745574790896713,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004430788667381035,
      "loss": 5.07,
      "mean_token_accuracy": 0.20311920195817948,
      "num_tokens": 52949881.0,
      "step": 30520
    },
    {
      "entropy": 5.696753454208374,
      "epoch": 2.374946508461389,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044306044000932374,
      "loss": 4.9257,
      "mean_token_accuracy": 0.20700712352991105,
      "num_tokens": 52957836.0,
      "step": 30525
    },
    {
      "entropy": 5.693720865249634,
      "epoch": 2.3753355378331062,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004430420107304943,
      "loss": 4.98,
      "mean_token_accuracy": 0.212174129486084,
      "num_tokens": 52965948.0,
      "step": 30530
    },
    {
      "entropy": 5.702986574172973,
      "epoch": 2.375724567204824,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044302357890189475,
      "loss": 4.96,
      "mean_token_accuracy": 0.20682018548250197,
      "num_tokens": 52974561.0,
      "step": 30535
    },
    {
      "entropy": 5.698385000228882,
      "epoch": 2.3761135965765416,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004430051445238049,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20959998071193695,
      "num_tokens": 52982918.0,
      "step": 30540
    },
    {
      "entropy": 5.765712881088257,
      "epoch": 2.376502625948259,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044298670759650426,
      "loss": 5.061,
      "mean_token_accuracy": 0.19887791574001312,
      "num_tokens": 52992119.0,
      "step": 30545
    },
    {
      "entropy": 5.620728540420532,
      "epoch": 2.3768916553199766,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004429682681202728,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21535825580358506,
      "num_tokens": 53000427.0,
      "step": 30550
    },
    {
      "entropy": 5.688116884231567,
      "epoch": 2.3772806846916943,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044294982609539027,
      "loss": 5.057,
      "mean_token_accuracy": 0.20019415020942688,
      "num_tokens": 53009615.0,
      "step": 30555
    },
    {
      "entropy": 5.744291877746582,
      "epoch": 2.377669714063412,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004429313815221363,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20610597878694534,
      "num_tokens": 53018357.0,
      "step": 30560
    },
    {
      "entropy": 5.70464015007019,
      "epoch": 2.3780587434351292,
      "grad_norm": 1.25,
      "learning_rate": 0.00044291293440079107,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2036253735423088,
      "num_tokens": 53027209.0,
      "step": 30565
    },
    {
      "entropy": 5.621303987503052,
      "epoch": 2.378447772806847,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004428944847316342,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2029971480369568,
      "num_tokens": 53035621.0,
      "step": 30570
    },
    {
      "entropy": 5.739584159851074,
      "epoch": 2.3788368021785646,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004428760325149458,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20876681953668594,
      "num_tokens": 53044249.0,
      "step": 30575
    },
    {
      "entropy": 5.732653713226318,
      "epoch": 2.379225831550282,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044285757775100584,
      "loss": 5.0313,
      "mean_token_accuracy": 0.2043532148003578,
      "num_tokens": 53052977.0,
      "step": 30580
    },
    {
      "entropy": 5.704536390304566,
      "epoch": 2.3796148609219996,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044283912044009436,
      "loss": 5.0196,
      "mean_token_accuracy": 0.20893745571374894,
      "num_tokens": 53061071.0,
      "step": 30585
    },
    {
      "entropy": 5.763770246505738,
      "epoch": 2.3800038902937173,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044282066058249123,
      "loss": 5.038,
      "mean_token_accuracy": 0.20631231963634492,
      "num_tokens": 53069237.0,
      "step": 30590
    },
    {
      "entropy": 5.7708981990814205,
      "epoch": 2.380392919665435,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004428021981784768,
      "loss": 5.0336,
      "mean_token_accuracy": 0.2021631419658661,
      "num_tokens": 53077991.0,
      "step": 30595
    },
    {
      "entropy": 5.675279903411865,
      "epoch": 2.3807819490371522,
      "grad_norm": 1.25,
      "learning_rate": 0.00044278373322833106,
      "loss": 4.9554,
      "mean_token_accuracy": 0.20584033578634262,
      "num_tokens": 53086263.0,
      "step": 30600
    },
    {
      "entropy": 5.669742918014526,
      "epoch": 2.38117097840887,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044276526573233416,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20648971498012542,
      "num_tokens": 53095048.0,
      "step": 30605
    },
    {
      "entropy": 5.789305830001831,
      "epoch": 2.381560007780587,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004427467956907664,
      "loss": 5.063,
      "mean_token_accuracy": 0.20561017990112304,
      "num_tokens": 53103199.0,
      "step": 30610
    },
    {
      "entropy": 5.804940032958984,
      "epoch": 2.381949037152305,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044272832310390814,
      "loss": 5.068,
      "mean_token_accuracy": 0.2033366948366165,
      "num_tokens": 53111677.0,
      "step": 30615
    },
    {
      "entropy": 5.737106132507324,
      "epoch": 2.3823380665240226,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004427098479720394,
      "loss": 5.0011,
      "mean_token_accuracy": 0.1996941789984703,
      "num_tokens": 53120136.0,
      "step": 30620
    },
    {
      "entropy": 5.762420415878296,
      "epoch": 2.3827270958957403,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044269137029544073,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20251599997282027,
      "num_tokens": 53128847.0,
      "step": 30625
    },
    {
      "entropy": 5.766170167922974,
      "epoch": 2.3831161252674575,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004426728900743924,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20086421221494674,
      "num_tokens": 53137307.0,
      "step": 30630
    },
    {
      "entropy": 5.69331727027893,
      "epoch": 2.3835051546391752,
      "grad_norm": 1.375,
      "learning_rate": 0.00044265440730917484,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20302721112966537,
      "num_tokens": 53145854.0,
      "step": 30635
    },
    {
      "entropy": 5.660968875885009,
      "epoch": 2.383894184010893,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044263592200006845,
      "loss": 5.0594,
      "mean_token_accuracy": 0.1970896378159523,
      "num_tokens": 53154779.0,
      "step": 30640
    },
    {
      "entropy": 5.750217294692993,
      "epoch": 2.38428321338261,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044261743414735383,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2013774737715721,
      "num_tokens": 53163344.0,
      "step": 30645
    },
    {
      "entropy": 5.7137984275817875,
      "epoch": 2.384672242754328,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004425989437513114,
      "loss": 4.9001,
      "mean_token_accuracy": 0.20931152999401093,
      "num_tokens": 53171989.0,
      "step": 30650
    },
    {
      "entropy": 5.726375770568848,
      "epoch": 2.3850612721260456,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004425804508122218,
      "loss": 5.0354,
      "mean_token_accuracy": 0.19894566684961318,
      "num_tokens": 53180612.0,
      "step": 30655
    },
    {
      "entropy": 5.715978574752808,
      "epoch": 2.3854503014977633,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044256195533036566,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20585851669311522,
      "num_tokens": 53188741.0,
      "step": 30660
    },
    {
      "entropy": 5.712654876708984,
      "epoch": 2.3858393308694805,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044254345730602345,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2066669210791588,
      "num_tokens": 53196857.0,
      "step": 30665
    },
    {
      "entropy": 5.678563785552979,
      "epoch": 2.386228360241198,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004425249567394759,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20422122329473497,
      "num_tokens": 53205264.0,
      "step": 30670
    },
    {
      "entropy": 5.7743183135986325,
      "epoch": 2.386617389612916,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044250645363100386,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21143311858177186,
      "num_tokens": 53213459.0,
      "step": 30675
    },
    {
      "entropy": 5.768850040435791,
      "epoch": 2.387006418984633,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044248794798088805,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19860921502113343,
      "num_tokens": 53221740.0,
      "step": 30680
    },
    {
      "entropy": 5.72217869758606,
      "epoch": 2.387395448356351,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004424694397894091,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2027123302221298,
      "num_tokens": 53230177.0,
      "step": 30685
    },
    {
      "entropy": 5.72634220123291,
      "epoch": 2.3877844777280686,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044245092905684794,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2028859406709671,
      "num_tokens": 53238238.0,
      "step": 30690
    },
    {
      "entropy": 5.769701528549194,
      "epoch": 2.3881735070997863,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044243241578348553,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2070941910147667,
      "num_tokens": 53246519.0,
      "step": 30695
    },
    {
      "entropy": 5.73446216583252,
      "epoch": 2.3885625364715035,
      "grad_norm": 1.125,
      "learning_rate": 0.0004424138999696027,
      "loss": 5.0699,
      "mean_token_accuracy": 0.1981464922428131,
      "num_tokens": 53255522.0,
      "step": 30700
    },
    {
      "entropy": 5.622249794006348,
      "epoch": 2.388951565843221,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004423953816154804,
      "loss": 4.9361,
      "mean_token_accuracy": 0.20958486646413804,
      "num_tokens": 53264414.0,
      "step": 30705
    },
    {
      "entropy": 5.721840238571167,
      "epoch": 2.389340595214939,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044237686072139967,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20067951679229737,
      "num_tokens": 53273403.0,
      "step": 30710
    },
    {
      "entropy": 5.808246040344239,
      "epoch": 2.389729624586656,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004423583372876414,
      "loss": 5.069,
      "mean_token_accuracy": 0.19770598262548447,
      "num_tokens": 53282361.0,
      "step": 30715
    },
    {
      "entropy": 5.7130931377410885,
      "epoch": 2.390118653958374,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044233981131448677,
      "loss": 4.9904,
      "mean_token_accuracy": 0.19889576137065887,
      "num_tokens": 53290890.0,
      "step": 30720
    },
    {
      "entropy": 5.608754873275757,
      "epoch": 2.3905076833300916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044232128280221683,
      "loss": 4.9158,
      "mean_token_accuracy": 0.2104830712080002,
      "num_tokens": 53299781.0,
      "step": 30725
    },
    {
      "entropy": 5.686108684539795,
      "epoch": 2.390896712701809,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004423027517511128,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20447646528482438,
      "num_tokens": 53308568.0,
      "step": 30730
    },
    {
      "entropy": 5.795271539688111,
      "epoch": 2.3912857420735265,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044228421816145573,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20579529851675032,
      "num_tokens": 53317310.0,
      "step": 30735
    },
    {
      "entropy": 5.7256128787994385,
      "epoch": 2.391674771445244,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044226568203352694,
      "loss": 4.9853,
      "mean_token_accuracy": 0.19957933574914932,
      "num_tokens": 53326022.0,
      "step": 30740
    },
    {
      "entropy": 5.74186749458313,
      "epoch": 2.3920638008169615,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044224714336760766,
      "loss": 4.9963,
      "mean_token_accuracy": 0.2039797604084015,
      "num_tokens": 53334090.0,
      "step": 30745
    },
    {
      "entropy": 5.671142482757569,
      "epoch": 2.392452830188679,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004422286021639792,
      "loss": 4.9422,
      "mean_token_accuracy": 0.1970006436109543,
      "num_tokens": 53342682.0,
      "step": 30750
    },
    {
      "entropy": 5.738464212417602,
      "epoch": 2.392841859560397,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004422100584229228,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20119790285825728,
      "num_tokens": 53350652.0,
      "step": 30755
    },
    {
      "entropy": 5.7039892196655275,
      "epoch": 2.3932308889321146,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044219151214472,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2025495231151581,
      "num_tokens": 53359339.0,
      "step": 30760
    },
    {
      "entropy": 5.705528640747071,
      "epoch": 2.393619918303832,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004421729633296521,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20618873238563537,
      "num_tokens": 53367617.0,
      "step": 30765
    },
    {
      "entropy": 5.7443849563598635,
      "epoch": 2.3940089476755495,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044215441197800054,
      "loss": 5.1268,
      "mean_token_accuracy": 0.1906303733587265,
      "num_tokens": 53376725.0,
      "step": 30770
    },
    {
      "entropy": 5.751028203964234,
      "epoch": 2.394397977047267,
      "grad_norm": 1.25,
      "learning_rate": 0.00044213585809004685,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20503025352954865,
      "num_tokens": 53384670.0,
      "step": 30775
    },
    {
      "entropy": 5.677694272994995,
      "epoch": 2.3947870064189845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004421173016660725,
      "loss": 4.9715,
      "mean_token_accuracy": 0.20712345242500305,
      "num_tokens": 53392944.0,
      "step": 30780
    },
    {
      "entropy": 5.70525541305542,
      "epoch": 2.395176035790702,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004420987427063592,
      "loss": 5.0625,
      "mean_token_accuracy": 0.19969112277030945,
      "num_tokens": 53401301.0,
      "step": 30785
    },
    {
      "entropy": 5.782529973983765,
      "epoch": 2.39556506516242,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044208018121118833,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20520539879798888,
      "num_tokens": 53409490.0,
      "step": 30790
    },
    {
      "entropy": 5.8067216873168945,
      "epoch": 2.3959540945341375,
      "grad_norm": 1.25,
      "learning_rate": 0.00044206161718084164,
      "loss": 5.1297,
      "mean_token_accuracy": 0.192756225168705,
      "num_tokens": 53418595.0,
      "step": 30795
    },
    {
      "entropy": 5.677145099639892,
      "epoch": 2.396343123905855,
      "grad_norm": 1.25,
      "learning_rate": 0.0004420430506156009,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21123088002204896,
      "num_tokens": 53427217.0,
      "step": 30800
    },
    {
      "entropy": 5.654741191864014,
      "epoch": 2.3967321532775725,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044202448151574764,
      "loss": 4.9396,
      "mean_token_accuracy": 0.211744225025177,
      "num_tokens": 53435875.0,
      "step": 30805
    },
    {
      "entropy": 5.734385871887207,
      "epoch": 2.39712118264929,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004420059098815638,
      "loss": 5.022,
      "mean_token_accuracy": 0.20412782281637193,
      "num_tokens": 53445466.0,
      "step": 30810
    },
    {
      "entropy": 5.652290534973145,
      "epoch": 2.3975102120210074,
      "grad_norm": 1.25,
      "learning_rate": 0.0004419873357133311,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2093501403927803,
      "num_tokens": 53453813.0,
      "step": 30815
    },
    {
      "entropy": 5.712499046325684,
      "epoch": 2.397899241392725,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004419687590113313,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20638117343187332,
      "num_tokens": 53462819.0,
      "step": 30820
    },
    {
      "entropy": 5.740939044952393,
      "epoch": 2.398288270764443,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044195017977584637,
      "loss": 5.0043,
      "mean_token_accuracy": 0.2060355454683304,
      "num_tokens": 53471297.0,
      "step": 30825
    },
    {
      "entropy": 5.737145805358887,
      "epoch": 2.39867730013616,
      "grad_norm": 1.25,
      "learning_rate": 0.00044193159800715823,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20364607572555543,
      "num_tokens": 53479473.0,
      "step": 30830
    },
    {
      "entropy": 5.686576795578003,
      "epoch": 2.399066329507878,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044191301370554874,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20454010665416716,
      "num_tokens": 53488542.0,
      "step": 30835
    },
    {
      "entropy": 5.714493370056152,
      "epoch": 2.3994553588795955,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044189442687129994,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21020812839269637,
      "num_tokens": 53497586.0,
      "step": 30840
    },
    {
      "entropy": 5.799126863479614,
      "epoch": 2.3998443882513127,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004418758375046939,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20898991078138351,
      "num_tokens": 53506368.0,
      "step": 30845
    },
    {
      "entropy": 5.712053442001343,
      "epoch": 2.4002334176230304,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044185724560601267,
      "loss": 4.9966,
      "mean_token_accuracy": 0.20110889673233032,
      "num_tokens": 53515735.0,
      "step": 30850
    },
    {
      "entropy": 5.758318614959717,
      "epoch": 2.400622446994748,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004418386511755382,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19416172206401824,
      "num_tokens": 53523780.0,
      "step": 30855
    },
    {
      "entropy": 5.7093486309051515,
      "epoch": 2.401011476366466,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004418200542135528,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20281765460968018,
      "num_tokens": 53532633.0,
      "step": 30860
    },
    {
      "entropy": 5.782886028289795,
      "epoch": 2.401400505738183,
      "grad_norm": 1.25,
      "learning_rate": 0.0004418014547203386,
      "loss": 5.0832,
      "mean_token_accuracy": 0.21037146002054213,
      "num_tokens": 53540767.0,
      "step": 30865
    },
    {
      "entropy": 5.736627626419067,
      "epoch": 2.401789535109901,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004417828526961778,
      "loss": 5.0602,
      "mean_token_accuracy": 0.2053644061088562,
      "num_tokens": 53548925.0,
      "step": 30870
    },
    {
      "entropy": 5.647260046005249,
      "epoch": 2.4021785644816185,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044176424814135266,
      "loss": 4.8321,
      "mean_token_accuracy": 0.22049898654222488,
      "num_tokens": 53557086.0,
      "step": 30875
    },
    {
      "entropy": 5.72654013633728,
      "epoch": 2.4025675938533357,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004417456410561455,
      "loss": 5.098,
      "mean_token_accuracy": 0.19393283426761626,
      "num_tokens": 53565898.0,
      "step": 30880
    },
    {
      "entropy": 5.696679735183716,
      "epoch": 2.4029566232250534,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004417270314408387,
      "loss": 5.068,
      "mean_token_accuracy": 0.19617422819137573,
      "num_tokens": 53574695.0,
      "step": 30885
    },
    {
      "entropy": 5.751157426834107,
      "epoch": 2.403345652596771,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044170841929571454,
      "loss": 5.0593,
      "mean_token_accuracy": 0.19871010780334472,
      "num_tokens": 53583854.0,
      "step": 30890
    },
    {
      "entropy": 5.708415508270264,
      "epoch": 2.403734681968489,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044168980462105543,
      "loss": 4.9896,
      "mean_token_accuracy": 0.21502301394939421,
      "num_tokens": 53593008.0,
      "step": 30895
    },
    {
      "entropy": 5.729823446273803,
      "epoch": 2.404123711340206,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004416711874171439,
      "loss": 4.9558,
      "mean_token_accuracy": 0.20564486384391784,
      "num_tokens": 53601548.0,
      "step": 30900
    },
    {
      "entropy": 5.782066297531128,
      "epoch": 2.404512740711924,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004416525676842624,
      "loss": 5.1352,
      "mean_token_accuracy": 0.19614185839891435,
      "num_tokens": 53609807.0,
      "step": 30905
    },
    {
      "entropy": 5.696769285202026,
      "epoch": 2.4049017700836415,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004416339454226933,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2162472650408745,
      "num_tokens": 53618879.0,
      "step": 30910
    },
    {
      "entropy": 5.690673351287842,
      "epoch": 2.4052907994553587,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004416153206327194,
      "loss": 4.9554,
      "mean_token_accuracy": 0.20473794490098954,
      "num_tokens": 53627318.0,
      "step": 30915
    },
    {
      "entropy": 5.6443909168243405,
      "epoch": 2.4056798288270764,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044159669331462326,
      "loss": 4.9772,
      "mean_token_accuracy": 0.20369113534688948,
      "num_tokens": 53635872.0,
      "step": 30920
    },
    {
      "entropy": 5.739658641815185,
      "epoch": 2.406068858198794,
      "grad_norm": 1.125,
      "learning_rate": 0.00044157806346868737,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20079739689826964,
      "num_tokens": 53644434.0,
      "step": 30925
    },
    {
      "entropy": 5.7133464336395265,
      "epoch": 2.406457887570512,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044155943109519454,
      "loss": 4.9499,
      "mean_token_accuracy": 0.20679675042629242,
      "num_tokens": 53652721.0,
      "step": 30930
    },
    {
      "entropy": 5.708706521987915,
      "epoch": 2.406846916942229,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004415407961944274,
      "loss": 5.001,
      "mean_token_accuracy": 0.2032891631126404,
      "num_tokens": 53660831.0,
      "step": 30935
    },
    {
      "entropy": 5.742891120910644,
      "epoch": 2.4072359463139468,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044152215876666883,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2122283697128296,
      "num_tokens": 53669335.0,
      "step": 30940
    },
    {
      "entropy": 5.760075807571411,
      "epoch": 2.407624975685664,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004415035188122016,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20741085559129716,
      "num_tokens": 53678004.0,
      "step": 30945
    },
    {
      "entropy": 5.731687259674072,
      "epoch": 2.4080140050573817,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044148487633130837,
      "loss": 4.9994,
      "mean_token_accuracy": 0.20332690924406052,
      "num_tokens": 53687300.0,
      "step": 30950
    },
    {
      "entropy": 5.673222398757934,
      "epoch": 2.4084030344290994,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044146623132427213,
      "loss": 4.932,
      "mean_token_accuracy": 0.2045376792550087,
      "num_tokens": 53695692.0,
      "step": 30955
    },
    {
      "entropy": 5.737061214447022,
      "epoch": 2.408792063800817,
      "grad_norm": 1.25,
      "learning_rate": 0.00044144758379137584,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2029051512479782,
      "num_tokens": 53704164.0,
      "step": 30960
    },
    {
      "entropy": 5.644395875930786,
      "epoch": 2.4091810931725344,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044142893373290236,
      "loss": 4.9478,
      "mean_token_accuracy": 0.20806066691875458,
      "num_tokens": 53712681.0,
      "step": 30965
    },
    {
      "entropy": 5.70630464553833,
      "epoch": 2.409570122544252,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004414102811491348,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2069537580013275,
      "num_tokens": 53721208.0,
      "step": 30970
    },
    {
      "entropy": 5.719575214385986,
      "epoch": 2.4099591519159698,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044139162604035597,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2037212610244751,
      "num_tokens": 53729590.0,
      "step": 30975
    },
    {
      "entropy": 5.716948223114014,
      "epoch": 2.410348181287687,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044137296840684917,
      "loss": 5.0378,
      "mean_token_accuracy": 0.19348290264606477,
      "num_tokens": 53737886.0,
      "step": 30980
    },
    {
      "entropy": 5.743378162384033,
      "epoch": 2.4107372106594047,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004413543082488973,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2056961849331856,
      "num_tokens": 53746397.0,
      "step": 30985
    },
    {
      "entropy": 5.701738452911377,
      "epoch": 2.4111262400311224,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004413356455667836,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19941812753677368,
      "num_tokens": 53755319.0,
      "step": 30990
    },
    {
      "entropy": 5.6213947296142575,
      "epoch": 2.41151526940284,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004413169803607913,
      "loss": 4.9792,
      "mean_token_accuracy": 0.20228902548551558,
      "num_tokens": 53763655.0,
      "step": 30995
    },
    {
      "entropy": 5.750853967666626,
      "epoch": 2.4119042987745574,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044129831263120344,
      "loss": 5.0016,
      "mean_token_accuracy": 0.20588153153657912,
      "num_tokens": 53772922.0,
      "step": 31000
    },
    {
      "entropy": 5.769703912734985,
      "epoch": 2.412293328146275,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004412796423783034,
      "loss": 4.9765,
      "mean_token_accuracy": 0.201437047123909,
      "num_tokens": 53781328.0,
      "step": 31005
    },
    {
      "entropy": 5.673709535598755,
      "epoch": 2.4126823575179928,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004412609696023745,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20982305854558944,
      "num_tokens": 53791025.0,
      "step": 31010
    },
    {
      "entropy": 5.714229583740234,
      "epoch": 2.41307138688971,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004412422943037,
      "loss": 5.0411,
      "mean_token_accuracy": 0.2015644818544388,
      "num_tokens": 53799423.0,
      "step": 31015
    },
    {
      "entropy": 5.730498170852661,
      "epoch": 2.4134604162614277,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044122361648256327,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20246349424123763,
      "num_tokens": 53808001.0,
      "step": 31020
    },
    {
      "entropy": 5.698866939544677,
      "epoch": 2.4138494456331454,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044120493613924766,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21059404015541078,
      "num_tokens": 53816354.0,
      "step": 31025
    },
    {
      "entropy": 5.673737478256226,
      "epoch": 2.414238475004863,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044118625327403677,
      "loss": 4.9367,
      "mean_token_accuracy": 0.2097831666469574,
      "num_tokens": 53825430.0,
      "step": 31030
    },
    {
      "entropy": 5.761926889419556,
      "epoch": 2.4146275043765804,
      "grad_norm": 1.25,
      "learning_rate": 0.000441167567887214,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20569602400064468,
      "num_tokens": 53833420.0,
      "step": 31035
    },
    {
      "entropy": 5.7541169166564945,
      "epoch": 2.415016533748298,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004411488799790629,
      "loss": 5.0777,
      "mean_token_accuracy": 0.1932936355471611,
      "num_tokens": 53841771.0,
      "step": 31040
    },
    {
      "entropy": 5.7707044124603275,
      "epoch": 2.4154055631200158,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044113018954986693,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2098231479525566,
      "num_tokens": 53850485.0,
      "step": 31045
    },
    {
      "entropy": 5.849239253997803,
      "epoch": 2.415794592491733,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004411114965999098,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19041448682546616,
      "num_tokens": 53859491.0,
      "step": 31050
    },
    {
      "entropy": 5.676351499557495,
      "epoch": 2.4161836218634507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000441092801129475,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2123714119195938,
      "num_tokens": 53868838.0,
      "step": 31055
    },
    {
      "entropy": 5.792420101165772,
      "epoch": 2.4165726512351684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044107410313884643,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19635931253433228,
      "num_tokens": 53877689.0,
      "step": 31060
    },
    {
      "entropy": 5.6513158798217775,
      "epoch": 2.4169616806068857,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044105540262830763,
      "loss": 4.8475,
      "mean_token_accuracy": 0.2141550973057747,
      "num_tokens": 53885769.0,
      "step": 31065
    },
    {
      "entropy": 5.604755401611328,
      "epoch": 2.4173507099786034,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004410366995981424,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2127706974744797,
      "num_tokens": 53894745.0,
      "step": 31070
    },
    {
      "entropy": 5.692781066894531,
      "epoch": 2.417739739350321,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044101799404863457,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20602525472640992,
      "num_tokens": 53903888.0,
      "step": 31075
    },
    {
      "entropy": 5.742588710784912,
      "epoch": 2.4181287687220383,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004409992859800679,
      "loss": 4.992,
      "mean_token_accuracy": 0.20943558663129808,
      "num_tokens": 53912422.0,
      "step": 31080
    },
    {
      "entropy": 5.696617937088012,
      "epoch": 2.418517798093756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004409805753927263,
      "loss": 4.9972,
      "mean_token_accuracy": 0.20264054983854293,
      "num_tokens": 53921181.0,
      "step": 31085
    },
    {
      "entropy": 5.7647092819213865,
      "epoch": 2.4189068274654737,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004409618622868936,
      "loss": 5.1111,
      "mean_token_accuracy": 0.2018714278936386,
      "num_tokens": 53929777.0,
      "step": 31090
    },
    {
      "entropy": 5.7654846668243405,
      "epoch": 2.4192958568371914,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044094314666285385,
      "loss": 5.0223,
      "mean_token_accuracy": 0.19341136515140533,
      "num_tokens": 53938678.0,
      "step": 31095
    },
    {
      "entropy": 5.746962022781372,
      "epoch": 2.4196848862089086,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044092442852089095,
      "loss": 4.9553,
      "mean_token_accuracy": 0.20984273850917817,
      "num_tokens": 53946158.0,
      "step": 31100
    },
    {
      "entropy": 5.732106351852417,
      "epoch": 2.4200739155806263,
      "grad_norm": 1.25,
      "learning_rate": 0.0004409057078612889,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20642614364624023,
      "num_tokens": 53954875.0,
      "step": 31105
    },
    {
      "entropy": 5.74470362663269,
      "epoch": 2.420462944952344,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044088698468433194,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20012181699275972,
      "num_tokens": 53963355.0,
      "step": 31110
    },
    {
      "entropy": 5.713160371780395,
      "epoch": 2.4208519743240613,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044086825899030383,
      "loss": 5.028,
      "mean_token_accuracy": 0.20461328327655792,
      "num_tokens": 53971861.0,
      "step": 31115
    },
    {
      "entropy": 5.736696672439575,
      "epoch": 2.421241003695779,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000440849530779489,
      "loss": 4.992,
      "mean_token_accuracy": 0.19765418618917466,
      "num_tokens": 53980415.0,
      "step": 31120
    },
    {
      "entropy": 5.78776593208313,
      "epoch": 2.4216300330674967,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004408308000521715,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20284083038568496,
      "num_tokens": 53989107.0,
      "step": 31125
    },
    {
      "entropy": 5.73125114440918,
      "epoch": 2.4220190624392144,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044081206680863556,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20722728073596955,
      "num_tokens": 53997938.0,
      "step": 31130
    },
    {
      "entropy": 5.694199895858764,
      "epoch": 2.4224080918109316,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044079333104916554,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21030941903591155,
      "num_tokens": 54006221.0,
      "step": 31135
    },
    {
      "entropy": 5.734976387023925,
      "epoch": 2.4227971211826493,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004407745927740454,
      "loss": 5.0247,
      "mean_token_accuracy": 0.2052517980337143,
      "num_tokens": 54014579.0,
      "step": 31140
    },
    {
      "entropy": 5.757489585876465,
      "epoch": 2.423186150554367,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004407558519835598,
      "loss": 4.9834,
      "mean_token_accuracy": 0.19983111917972565,
      "num_tokens": 54024119.0,
      "step": 31145
    },
    {
      "entropy": 5.816200113296508,
      "epoch": 2.4235751799260843,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044073710867799295,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19989494681358339,
      "num_tokens": 54032548.0,
      "step": 31150
    },
    {
      "entropy": 5.721306037902832,
      "epoch": 2.423964209297802,
      "grad_norm": 1.25,
      "learning_rate": 0.0004407183628576293,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2098286807537079,
      "num_tokens": 54041210.0,
      "step": 31155
    },
    {
      "entropy": 5.711873006820679,
      "epoch": 2.4243532386695197,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004406996145227532,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20778978168964385,
      "num_tokens": 54050056.0,
      "step": 31160
    },
    {
      "entropy": 5.741904306411743,
      "epoch": 2.424742268041237,
      "grad_norm": 1.125,
      "learning_rate": 0.0004406808636736492,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2054418757557869,
      "num_tokens": 54058752.0,
      "step": 31165
    },
    {
      "entropy": 5.682488632202149,
      "epoch": 2.4251312974129546,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044066211031060183,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21279467940330504,
      "num_tokens": 54067496.0,
      "step": 31170
    },
    {
      "entropy": 5.6209217548370365,
      "epoch": 2.4255203267846723,
      "grad_norm": 1.25,
      "learning_rate": 0.0004406433544338956,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2060035765171051,
      "num_tokens": 54075832.0,
      "step": 31175
    },
    {
      "entropy": 5.8048523426055905,
      "epoch": 2.4259093561563896,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004406245960438151,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19790201783180236,
      "num_tokens": 54084642.0,
      "step": 31180
    },
    {
      "entropy": 5.704027843475342,
      "epoch": 2.4262983855281073,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000440605835140645,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21370572745800018,
      "num_tokens": 54092759.0,
      "step": 31185
    },
    {
      "entropy": 5.764912796020508,
      "epoch": 2.426687414899825,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004405870717246699,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20615483969449996,
      "num_tokens": 54101413.0,
      "step": 31190
    },
    {
      "entropy": 5.779792451858521,
      "epoch": 2.4270764442715427,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044056830579617464,
      "loss": 5.0536,
      "mean_token_accuracy": 0.19994451999664306,
      "num_tokens": 54110237.0,
      "step": 31195
    },
    {
      "entropy": 5.773042154312134,
      "epoch": 2.42746547364326,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044054953735544376,
      "loss": 5.1086,
      "mean_token_accuracy": 0.2025374412536621,
      "num_tokens": 54119103.0,
      "step": 31200
    },
    {
      "entropy": 5.72314224243164,
      "epoch": 2.4278545030149776,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004405307664027622,
      "loss": 5.012,
      "mean_token_accuracy": 0.20574668198823928,
      "num_tokens": 54127696.0,
      "step": 31205
    },
    {
      "entropy": 5.6964170932769775,
      "epoch": 2.4282435323866953,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004405119929384147,
      "loss": 4.9145,
      "mean_token_accuracy": 0.2153623178601265,
      "num_tokens": 54135607.0,
      "step": 31210
    },
    {
      "entropy": 5.72726263999939,
      "epoch": 2.4286325617584126,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004404932169626862,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20313965380191804,
      "num_tokens": 54144147.0,
      "step": 31215
    },
    {
      "entropy": 5.691176795959473,
      "epoch": 2.4290215911301303,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004404744384758615,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2086134746670723,
      "num_tokens": 54152010.0,
      "step": 31220
    },
    {
      "entropy": 5.688778877258301,
      "epoch": 2.429410620501848,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004404556574782256,
      "loss": 5.0729,
      "mean_token_accuracy": 0.21265822052955627,
      "num_tokens": 54160923.0,
      "step": 31225
    },
    {
      "entropy": 5.685829734802246,
      "epoch": 2.4297996498735657,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044043687397006336,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20167548209428787,
      "num_tokens": 54169688.0,
      "step": 31230
    },
    {
      "entropy": 5.751639413833618,
      "epoch": 2.430188679245283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044041808795166,
      "loss": 4.9973,
      "mean_token_accuracy": 0.20408645570278167,
      "num_tokens": 54178423.0,
      "step": 31235
    },
    {
      "entropy": 5.706020069122315,
      "epoch": 2.4305777086170006,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004403992994233004,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20605299472808838,
      "num_tokens": 54186939.0,
      "step": 31240
    },
    {
      "entropy": 5.721589517593384,
      "epoch": 2.4309667379887183,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044038050838526974,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20286224335432052,
      "num_tokens": 54195758.0,
      "step": 31245
    },
    {
      "entropy": 5.712616777420044,
      "epoch": 2.4313557673604356,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000440361714837853,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21455978602170944,
      "num_tokens": 54205139.0,
      "step": 31250
    },
    {
      "entropy": 5.712167167663575,
      "epoch": 2.4317447967321533,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004403429187813355,
      "loss": 4.949,
      "mean_token_accuracy": 0.20790459960699081,
      "num_tokens": 54213346.0,
      "step": 31255
    },
    {
      "entropy": 5.688876438140869,
      "epoch": 2.432133826103871,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004403241202160024,
      "loss": 4.8904,
      "mean_token_accuracy": 0.21033211201429367,
      "num_tokens": 54221796.0,
      "step": 31260
    },
    {
      "entropy": 5.764972400665283,
      "epoch": 2.432522855475588,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004403053191421388,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19257381856441497,
      "num_tokens": 54230166.0,
      "step": 31265
    },
    {
      "entropy": 5.6980335235595705,
      "epoch": 2.432911884847306,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004402865155600302,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20515516847372056,
      "num_tokens": 54238986.0,
      "step": 31270
    },
    {
      "entropy": 5.758154630661011,
      "epoch": 2.4333009142190236,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044026770946996183,
      "loss": 4.966,
      "mean_token_accuracy": 0.2122556447982788,
      "num_tokens": 54248178.0,
      "step": 31275
    },
    {
      "entropy": 5.668928003311157,
      "epoch": 2.433689943590741,
      "grad_norm": 1.25,
      "learning_rate": 0.00044024890087221896,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2069834738969803,
      "num_tokens": 54256320.0,
      "step": 31280
    },
    {
      "entropy": 5.669818544387818,
      "epoch": 2.4340789729624586,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044023008976708705,
      "loss": 4.9376,
      "mean_token_accuracy": 0.21235679537057878,
      "num_tokens": 54264209.0,
      "step": 31285
    },
    {
      "entropy": 5.665422248840332,
      "epoch": 2.4344680023341763,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004402112761548515,
      "loss": 4.8726,
      "mean_token_accuracy": 0.21417779177427293,
      "num_tokens": 54271943.0,
      "step": 31290
    },
    {
      "entropy": 5.684609317779541,
      "epoch": 2.434857031705894,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044019246003579776,
      "loss": 5.0169,
      "mean_token_accuracy": 0.19852292984724046,
      "num_tokens": 54280524.0,
      "step": 31295
    },
    {
      "entropy": 5.735909414291382,
      "epoch": 2.435246061077611,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004401736414102115,
      "loss": 5.0001,
      "mean_token_accuracy": 0.20426616072654724,
      "num_tokens": 54289479.0,
      "step": 31300
    },
    {
      "entropy": 5.747764682769775,
      "epoch": 2.435635090449329,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000440154820278378,
      "loss": 4.9918,
      "mean_token_accuracy": 0.20194064527750016,
      "num_tokens": 54297502.0,
      "step": 31305
    },
    {
      "entropy": 5.759793043136597,
      "epoch": 2.4360241198210466,
      "grad_norm": 1.25,
      "learning_rate": 0.0004401359966405831,
      "loss": 5.0445,
      "mean_token_accuracy": 0.19651620984077453,
      "num_tokens": 54305691.0,
      "step": 31310
    },
    {
      "entropy": 5.8079344749450685,
      "epoch": 2.436413149192764,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044011717049711216,
      "loss": 5.0599,
      "mean_token_accuracy": 0.1966112360358238,
      "num_tokens": 54315582.0,
      "step": 31315
    },
    {
      "entropy": 5.7623992443084715,
      "epoch": 2.4368021785644816,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044009834184825103,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20213448405265808,
      "num_tokens": 54324102.0,
      "step": 31320
    },
    {
      "entropy": 5.790055561065674,
      "epoch": 2.4371912079361993,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004400795106942853,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1996346414089203,
      "num_tokens": 54332618.0,
      "step": 31325
    },
    {
      "entropy": 5.786337089538574,
      "epoch": 2.437580237307917,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044006067703550083,
      "loss": 5.027,
      "mean_token_accuracy": 0.20216328054666519,
      "num_tokens": 54341236.0,
      "step": 31330
    },
    {
      "entropy": 5.686537981033325,
      "epoch": 2.437969266679634,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004400418408721833,
      "loss": 4.9212,
      "mean_token_accuracy": 0.21061649322509765,
      "num_tokens": 54349817.0,
      "step": 31335
    },
    {
      "entropy": 5.781250524520874,
      "epoch": 2.438358296051352,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044002300220461843,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2016315445303917,
      "num_tokens": 54358767.0,
      "step": 31340
    },
    {
      "entropy": 5.648321914672851,
      "epoch": 2.4387473254230696,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004400041610330922,
      "loss": 4.902,
      "mean_token_accuracy": 0.21588591486215591,
      "num_tokens": 54367208.0,
      "step": 31345
    },
    {
      "entropy": 5.698946094512939,
      "epoch": 2.439136354794787,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004399853173578905,
      "loss": 5.019,
      "mean_token_accuracy": 0.20674995183944703,
      "num_tokens": 54376234.0,
      "step": 31350
    },
    {
      "entropy": 5.723383569717408,
      "epoch": 2.4395253841665046,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043996647117929916,
      "loss": 5.0296,
      "mean_token_accuracy": 0.2062743604183197,
      "num_tokens": 54384828.0,
      "step": 31355
    },
    {
      "entropy": 5.750326299667359,
      "epoch": 2.4399144135382222,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004399476224976043,
      "loss": 5.0438,
      "mean_token_accuracy": 0.20403137654066086,
      "num_tokens": 54394068.0,
      "step": 31360
    },
    {
      "entropy": 5.722454643249511,
      "epoch": 2.44030344290994,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043992877131309165,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20534244030714036,
      "num_tokens": 54403667.0,
      "step": 31365
    },
    {
      "entropy": 5.702554178237915,
      "epoch": 2.440692472281657,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004399099176260475,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2132360354065895,
      "num_tokens": 54411614.0,
      "step": 31370
    },
    {
      "entropy": 5.64546480178833,
      "epoch": 2.441081501653375,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004398910614367579,
      "loss": 4.9217,
      "mean_token_accuracy": 0.2075148731470108,
      "num_tokens": 54420084.0,
      "step": 31375
    },
    {
      "entropy": 5.66742730140686,
      "epoch": 2.441470531025092,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043987220274550876,
      "loss": 5.032,
      "mean_token_accuracy": 0.2055381268262863,
      "num_tokens": 54429100.0,
      "step": 31380
    },
    {
      "entropy": 5.795994091033935,
      "epoch": 2.44185956039681,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043985334155258643,
      "loss": 5.0707,
      "mean_token_accuracy": 0.2022236928343773,
      "num_tokens": 54437900.0,
      "step": 31385
    },
    {
      "entropy": 5.751574945449829,
      "epoch": 2.4422485897685275,
      "grad_norm": 1.203125,
      "learning_rate": 0.000439834477858277,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2022995486855507,
      "num_tokens": 54445840.0,
      "step": 31390
    },
    {
      "entropy": 5.686687707901001,
      "epoch": 2.4426376191402452,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043981561166286676,
      "loss": 5.0286,
      "mean_token_accuracy": 0.20303537845611572,
      "num_tokens": 54454116.0,
      "step": 31395
    },
    {
      "entropy": 5.6905069828033445,
      "epoch": 2.4430266485119625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000439796742966642,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20729911625385283,
      "num_tokens": 54462787.0,
      "step": 31400
    },
    {
      "entropy": 5.708018207550049,
      "epoch": 2.44341567788368,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043977787176988884,
      "loss": 5.0273,
      "mean_token_accuracy": 0.19933186322450638,
      "num_tokens": 54472706.0,
      "step": 31405
    },
    {
      "entropy": 5.751232576370239,
      "epoch": 2.443804707255398,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004397589980728938,
      "loss": 4.9997,
      "mean_token_accuracy": 0.200349822640419,
      "num_tokens": 54482016.0,
      "step": 31410
    },
    {
      "entropy": 5.693514442443847,
      "epoch": 2.444193736627115,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043974012187594324,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21324117183685304,
      "num_tokens": 54489943.0,
      "step": 31415
    },
    {
      "entropy": 5.724748754501343,
      "epoch": 2.444582765998833,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004397212431793235,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20083846747875214,
      "num_tokens": 54498251.0,
      "step": 31420
    },
    {
      "entropy": 5.628053522109985,
      "epoch": 2.4449717953705505,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043970236198332107,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21194072365760802,
      "num_tokens": 54506862.0,
      "step": 31425
    },
    {
      "entropy": 5.662540578842163,
      "epoch": 2.4453608247422682,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043968347828822246,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21807021498680115,
      "num_tokens": 54515207.0,
      "step": 31430
    },
    {
      "entropy": 5.745637273788452,
      "epoch": 2.4457498541139855,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004396645920943142,
      "loss": 5.0564,
      "mean_token_accuracy": 0.2068910852074623,
      "num_tokens": 54523643.0,
      "step": 31435
    },
    {
      "entropy": 5.651769208908081,
      "epoch": 2.446138883485703,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004396457034018828,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22324295938014985,
      "num_tokens": 54532805.0,
      "step": 31440
    },
    {
      "entropy": 5.768929052352905,
      "epoch": 2.446527912857421,
      "grad_norm": 1.328125,
      "learning_rate": 0.000439626812211215,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19941968768835067,
      "num_tokens": 54541525.0,
      "step": 31445
    },
    {
      "entropy": 5.734343957901001,
      "epoch": 2.446916942229138,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043960791852259735,
      "loss": 5.0648,
      "mean_token_accuracy": 0.198200823366642,
      "num_tokens": 54550034.0,
      "step": 31450
    },
    {
      "entropy": 5.764940929412842,
      "epoch": 2.447305971600856,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004395890223363165,
      "loss": 4.973,
      "mean_token_accuracy": 0.20376267433166503,
      "num_tokens": 54559498.0,
      "step": 31455
    },
    {
      "entropy": 5.730062007904053,
      "epoch": 2.4476950009725735,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043957012365265915,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20903342068195344,
      "num_tokens": 54568397.0,
      "step": 31460
    },
    {
      "entropy": 5.713685178756714,
      "epoch": 2.4480840303442912,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043955122247191214,
      "loss": 5.0759,
      "mean_token_accuracy": 0.1958493933081627,
      "num_tokens": 54576735.0,
      "step": 31465
    },
    {
      "entropy": 5.79483699798584,
      "epoch": 2.4484730597160085,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004395323187943623,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20244448632001877,
      "num_tokens": 54586265.0,
      "step": 31470
    },
    {
      "entropy": 5.796083974838257,
      "epoch": 2.448862089087726,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004395134126202964,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2064693659543991,
      "num_tokens": 54595307.0,
      "step": 31475
    },
    {
      "entropy": 5.6944561958312985,
      "epoch": 2.449251118459444,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004394945039500012,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2051499828696251,
      "num_tokens": 54603886.0,
      "step": 31480
    },
    {
      "entropy": 5.7135094165802,
      "epoch": 2.449640147831161,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043947559278376385,
      "loss": 5.052,
      "mean_token_accuracy": 0.2023889347910881,
      "num_tokens": 54613165.0,
      "step": 31485
    },
    {
      "entropy": 5.785331726074219,
      "epoch": 2.450029177202879,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043945667912187117,
      "loss": 4.9578,
      "mean_token_accuracy": 0.20812605172395707,
      "num_tokens": 54622226.0,
      "step": 31490
    },
    {
      "entropy": 5.678793811798096,
      "epoch": 2.4504182065745965,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004394377629646101,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2046058416366577,
      "num_tokens": 54631389.0,
      "step": 31495
    },
    {
      "entropy": 5.696940994262695,
      "epoch": 2.4508072359463138,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043941884431226763,
      "loss": 4.9853,
      "mean_token_accuracy": 0.20398652851581572,
      "num_tokens": 54640033.0,
      "step": 31500
    },
    {
      "entropy": 5.703508996963501,
      "epoch": 2.4511962653180315,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000439399923165131,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20419330447912215,
      "num_tokens": 54648828.0,
      "step": 31505
    },
    {
      "entropy": 5.687992668151855,
      "epoch": 2.451585294689749,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004393809995234872,
      "loss": 5.0,
      "mean_token_accuracy": 0.20139135867357255,
      "num_tokens": 54657132.0,
      "step": 31510
    },
    {
      "entropy": 5.733045673370361,
      "epoch": 2.4519743240614664,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004393620733876233,
      "loss": 4.9715,
      "mean_token_accuracy": 0.20381092727184297,
      "num_tokens": 54666261.0,
      "step": 31515
    },
    {
      "entropy": 5.653453683853149,
      "epoch": 2.452363353433184,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004393431447578267,
      "loss": 4.9563,
      "mean_token_accuracy": 0.20811671614646912,
      "num_tokens": 54675200.0,
      "step": 31520
    },
    {
      "entropy": 5.681611490249634,
      "epoch": 2.452752382804902,
      "grad_norm": 1.25,
      "learning_rate": 0.00043932421363438427,
      "loss": 4.954,
      "mean_token_accuracy": 0.20720795840024947,
      "num_tokens": 54683659.0,
      "step": 31525
    },
    {
      "entropy": 5.624059963226318,
      "epoch": 2.4531414121766195,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004393052800175835,
      "loss": 4.8811,
      "mean_token_accuracy": 0.21455943137407302,
      "num_tokens": 54691903.0,
      "step": 31530
    },
    {
      "entropy": 5.612323188781739,
      "epoch": 2.4535304415483368,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043928634390771166,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20599865466356276,
      "num_tokens": 54700512.0,
      "step": 31535
    },
    {
      "entropy": 5.695890665054321,
      "epoch": 2.4539194709200545,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043926740530505603,
      "loss": 4.964,
      "mean_token_accuracy": 0.2079227924346924,
      "num_tokens": 54709144.0,
      "step": 31540
    },
    {
      "entropy": 5.736378192901611,
      "epoch": 2.454308500291772,
      "grad_norm": 1.265625,
      "learning_rate": 0.000439248464209904,
      "loss": 5.001,
      "mean_token_accuracy": 0.20886046886444093,
      "num_tokens": 54717733.0,
      "step": 31545
    },
    {
      "entropy": 5.694670534133911,
      "epoch": 2.4546975296634894,
      "grad_norm": 1.25,
      "learning_rate": 0.000439229520622543,
      "loss": 4.962,
      "mean_token_accuracy": 0.2138854131102562,
      "num_tokens": 54726120.0,
      "step": 31550
    },
    {
      "entropy": 5.692197227478028,
      "epoch": 2.455086559035207,
      "grad_norm": 1.125,
      "learning_rate": 0.0004392105745432603,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21446151435375213,
      "num_tokens": 54735639.0,
      "step": 31555
    },
    {
      "entropy": 5.731448030471801,
      "epoch": 2.455475588406925,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004391916259723436,
      "loss": 4.9313,
      "mean_token_accuracy": 0.20501159876585007,
      "num_tokens": 54743928.0,
      "step": 31560
    },
    {
      "entropy": 5.74853253364563,
      "epoch": 2.4558646177786425,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043917267491008023,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20648407936096191,
      "num_tokens": 54752409.0,
      "step": 31565
    },
    {
      "entropy": 5.716746377944946,
      "epoch": 2.4562536471503598,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004391537213567579,
      "loss": 5.0308,
      "mean_token_accuracy": 0.20462253242731093,
      "num_tokens": 54761018.0,
      "step": 31570
    },
    {
      "entropy": 5.717392921447754,
      "epoch": 2.4566426765220775,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004391347653126641,
      "loss": 4.9554,
      "mean_token_accuracy": 0.20743818134069442,
      "num_tokens": 54769479.0,
      "step": 31575
    },
    {
      "entropy": 5.620333814620972,
      "epoch": 2.457031705893795,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043911580677808646,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21204301416873933,
      "num_tokens": 54778310.0,
      "step": 31580
    },
    {
      "entropy": 5.708395385742188,
      "epoch": 2.4574207352655124,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043909684575331264,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21367682367563248,
      "num_tokens": 54788237.0,
      "step": 31585
    },
    {
      "entropy": 5.697439956665039,
      "epoch": 2.45780976463723,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004390778822386304,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20962996035814285,
      "num_tokens": 54796746.0,
      "step": 31590
    },
    {
      "entropy": 5.643086051940918,
      "epoch": 2.458198794008948,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043905891623432754,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20551430135965348,
      "num_tokens": 54805572.0,
      "step": 31595
    },
    {
      "entropy": 5.699038171768189,
      "epoch": 2.458587823380665,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043903994774069165,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20952487289905547,
      "num_tokens": 54814631.0,
      "step": 31600
    },
    {
      "entropy": 5.792125701904297,
      "epoch": 2.4589768527523828,
      "grad_norm": 1.25,
      "learning_rate": 0.0004390209767580107,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20493804067373275,
      "num_tokens": 54822753.0,
      "step": 31605
    },
    {
      "entropy": 5.718414258956909,
      "epoch": 2.4593658821241005,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004390020032865725,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21025582402944565,
      "num_tokens": 54830969.0,
      "step": 31610
    },
    {
      "entropy": 5.650636482238769,
      "epoch": 2.4597549114958177,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004389830273266649,
      "loss": 5.0088,
      "mean_token_accuracy": 0.20762539952993392,
      "num_tokens": 54840295.0,
      "step": 31615
    },
    {
      "entropy": 5.841305112838745,
      "epoch": 2.4601439408675354,
      "grad_norm": 1.40625,
      "learning_rate": 0.00043896404887857596,
      "loss": 5.0927,
      "mean_token_accuracy": 0.20153098553419113,
      "num_tokens": 54848583.0,
      "step": 31620
    },
    {
      "entropy": 5.744894742965698,
      "epoch": 2.460532970239253,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004389450679425935,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21357618421316146,
      "num_tokens": 54857450.0,
      "step": 31625
    },
    {
      "entropy": 5.7445775985717775,
      "epoch": 2.460921999610971,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004389260845190055,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20098925679922103,
      "num_tokens": 54865928.0,
      "step": 31630
    },
    {
      "entropy": 5.761274290084839,
      "epoch": 2.461311028982688,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043890709860810014,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2015048936009407,
      "num_tokens": 54874983.0,
      "step": 31635
    },
    {
      "entropy": 5.72711296081543,
      "epoch": 2.4617000583544058,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043888811021016545,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20309460163116455,
      "num_tokens": 54883488.0,
      "step": 31640
    },
    {
      "entropy": 5.639499664306641,
      "epoch": 2.4620890877261234,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043886911932548953,
      "loss": 4.9713,
      "mean_token_accuracy": 0.20747045129537584,
      "num_tokens": 54892865.0,
      "step": 31645
    },
    {
      "entropy": 5.731367206573486,
      "epoch": 2.4624781170978407,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004388501259543605,
      "loss": 5.056,
      "mean_token_accuracy": 0.19783622473478318,
      "num_tokens": 54901714.0,
      "step": 31650
    },
    {
      "entropy": 5.663808584213257,
      "epoch": 2.4628671464695584,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004388311300970667,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20190518349409103,
      "num_tokens": 54910912.0,
      "step": 31655
    },
    {
      "entropy": 5.781589603424072,
      "epoch": 2.463256175841276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004388121317538962,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19187508374452591,
      "num_tokens": 54919613.0,
      "step": 31660
    },
    {
      "entropy": 5.74063286781311,
      "epoch": 2.463645205212994,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004387931309251374,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2064862906932831,
      "num_tokens": 54928794.0,
      "step": 31665
    },
    {
      "entropy": 5.656760358810425,
      "epoch": 2.464034234584711,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004387741276110785,
      "loss": 4.9272,
      "mean_token_accuracy": 0.20883956998586656,
      "num_tokens": 54937826.0,
      "step": 31670
    },
    {
      "entropy": 5.7515538215637205,
      "epoch": 2.4644232639564287,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043875512181200784,
      "loss": 5.006,
      "mean_token_accuracy": 0.20132196247577666,
      "num_tokens": 54946541.0,
      "step": 31675
    },
    {
      "entropy": 5.773842906951904,
      "epoch": 2.4648122933281464,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004387361135282139,
      "loss": 5.021,
      "mean_token_accuracy": 0.20059731900691985,
      "num_tokens": 54954540.0,
      "step": 31680
    },
    {
      "entropy": 5.689413785934448,
      "epoch": 2.4652013226998637,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043871710275998507,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2037371963262558,
      "num_tokens": 54962784.0,
      "step": 31685
    },
    {
      "entropy": 5.664430427551269,
      "epoch": 2.4655903520715814,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004386980895076097,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21631023287773132,
      "num_tokens": 54971572.0,
      "step": 31690
    },
    {
      "entropy": 5.708453226089477,
      "epoch": 2.465979381443299,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043867907377137646,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21270301192998886,
      "num_tokens": 54980161.0,
      "step": 31695
    },
    {
      "entropy": 5.736630821228028,
      "epoch": 2.4663684108150163,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043866005555157376,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20716388374567032,
      "num_tokens": 54988219.0,
      "step": 31700
    },
    {
      "entropy": 5.731163024902344,
      "epoch": 2.466757440186734,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004386410348484903,
      "loss": 4.9257,
      "mean_token_accuracy": 0.20828355252742767,
      "num_tokens": 54996550.0,
      "step": 31705
    },
    {
      "entropy": 5.740950441360473,
      "epoch": 2.4671464695584517,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004386220116624145,
      "loss": 5.0744,
      "mean_token_accuracy": 0.20379638075828552,
      "num_tokens": 55005269.0,
      "step": 31710
    },
    {
      "entropy": 5.722807598114014,
      "epoch": 2.467535498930169,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043860298599363506,
      "loss": 4.8927,
      "mean_token_accuracy": 0.21543223261833191,
      "num_tokens": 55014264.0,
      "step": 31715
    },
    {
      "entropy": 5.701769208908081,
      "epoch": 2.4679245283018867,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004385839578424408,
      "loss": 4.9556,
      "mean_token_accuracy": 0.20888547897338866,
      "num_tokens": 55023597.0,
      "step": 31720
    },
    {
      "entropy": 5.713161993026733,
      "epoch": 2.4683135576736044,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043856492720912037,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20741692930459976,
      "num_tokens": 55032860.0,
      "step": 31725
    },
    {
      "entropy": 5.734894227981568,
      "epoch": 2.468702587045322,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004385458940939624,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20935723483562468,
      "num_tokens": 55041275.0,
      "step": 31730
    },
    {
      "entropy": 5.749799537658691,
      "epoch": 2.4690916164170393,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004385268584972559,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20238760411739348,
      "num_tokens": 55050871.0,
      "step": 31735
    },
    {
      "entropy": 5.720091867446899,
      "epoch": 2.469480645788757,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004385078204192896,
      "loss": 4.9949,
      "mean_token_accuracy": 0.19972536712884903,
      "num_tokens": 55060353.0,
      "step": 31740
    },
    {
      "entropy": 5.701060104370117,
      "epoch": 2.4698696751604747,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004384887798603523,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20290753692388536,
      "num_tokens": 55069002.0,
      "step": 31745
    },
    {
      "entropy": 5.758010673522949,
      "epoch": 2.470258704532192,
      "grad_norm": 1.265625,
      "learning_rate": 0.000438469736820733,
      "loss": 5.0441,
      "mean_token_accuracy": 0.19967294931411744,
      "num_tokens": 55078369.0,
      "step": 31750
    },
    {
      "entropy": 5.824708127975464,
      "epoch": 2.4706477339039097,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043845069130072067,
      "loss": 5.1065,
      "mean_token_accuracy": 0.19176014214754106,
      "num_tokens": 55087597.0,
      "step": 31755
    },
    {
      "entropy": 5.679225158691406,
      "epoch": 2.4710367632756274,
      "grad_norm": 1.25,
      "learning_rate": 0.0004384316433006043,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21124570369720458,
      "num_tokens": 55095906.0,
      "step": 31760
    },
    {
      "entropy": 5.661124849319458,
      "epoch": 2.471425792647345,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043841259282067276,
      "loss": 4.9551,
      "mean_token_accuracy": 0.20399525612592698,
      "num_tokens": 55104401.0,
      "step": 31765
    },
    {
      "entropy": 5.6653076171875,
      "epoch": 2.4718148220190623,
      "grad_norm": 1.125,
      "learning_rate": 0.00043839353986121533,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2045558735728264,
      "num_tokens": 55113075.0,
      "step": 31770
    },
    {
      "entropy": 5.694813394546509,
      "epoch": 2.47220385139078,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004383744844225209,
      "loss": 4.9241,
      "mean_token_accuracy": 0.20841654092073442,
      "num_tokens": 55122173.0,
      "step": 31775
    },
    {
      "entropy": 5.742166137695312,
      "epoch": 2.4725928807624977,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043835542650487875,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20289955884218216,
      "num_tokens": 55130356.0,
      "step": 31780
    },
    {
      "entropy": 5.665129518508911,
      "epoch": 2.472981910134215,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004383363661085781,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2054634690284729,
      "num_tokens": 55138985.0,
      "step": 31785
    },
    {
      "entropy": 5.677022695541382,
      "epoch": 2.4733709395059327,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004383173032339079,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2069918245077133,
      "num_tokens": 55148053.0,
      "step": 31790
    },
    {
      "entropy": 5.805190563201904,
      "epoch": 2.4737599688776504,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004382982378811577,
      "loss": 5.1007,
      "mean_token_accuracy": 0.19765211641788483,
      "num_tokens": 55156815.0,
      "step": 31795
    },
    {
      "entropy": 5.671144437789917,
      "epoch": 2.474148998249368,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004382791700506166,
      "loss": 4.8661,
      "mean_token_accuracy": 0.21918827295303345,
      "num_tokens": 55165007.0,
      "step": 31800
    },
    {
      "entropy": 5.705432653427124,
      "epoch": 2.4745380276210853,
      "grad_norm": 1.234375,
      "learning_rate": 0.000438260099742574,
      "loss": 4.9988,
      "mean_token_accuracy": 0.2077239990234375,
      "num_tokens": 55173784.0,
      "step": 31805
    },
    {
      "entropy": 5.695836782455444,
      "epoch": 2.474927056992803,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004382410269573192,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20612655729055404,
      "num_tokens": 55182436.0,
      "step": 31810
    },
    {
      "entropy": 5.754938554763794,
      "epoch": 2.4753160863645203,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043822195169514167,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21072992831468582,
      "num_tokens": 55190083.0,
      "step": 31815
    },
    {
      "entropy": 5.727593278884887,
      "epoch": 2.475705115736238,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043820287395633086,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19770146906375885,
      "num_tokens": 55200154.0,
      "step": 31820
    },
    {
      "entropy": 5.775181102752685,
      "epoch": 2.4760941451079557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004381837937411761,
      "loss": 5.005,
      "mean_token_accuracy": 0.20843105316162108,
      "num_tokens": 55209672.0,
      "step": 31825
    },
    {
      "entropy": 5.787992906570435,
      "epoch": 2.4764831744796734,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043816471104996706,
      "loss": 5.0719,
      "mean_token_accuracy": 0.20008490681648256,
      "num_tokens": 55218362.0,
      "step": 31830
    },
    {
      "entropy": 5.704986143112182,
      "epoch": 2.4768722038513906,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004381456258829933,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2137981414794922,
      "num_tokens": 55226834.0,
      "step": 31835
    },
    {
      "entropy": 5.636142253875732,
      "epoch": 2.4772612332231083,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043812653824054426,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2027864158153534,
      "num_tokens": 55236325.0,
      "step": 31840
    },
    {
      "entropy": 5.730320644378662,
      "epoch": 2.477650262594826,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004381074481229097,
      "loss": 4.9575,
      "mean_token_accuracy": 0.2067609816789627,
      "num_tokens": 55245007.0,
      "step": 31845
    },
    {
      "entropy": 5.727527379989624,
      "epoch": 2.4780392919665433,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004380883555303791,
      "loss": 4.918,
      "mean_token_accuracy": 0.21145955771207808,
      "num_tokens": 55253967.0,
      "step": 31850
    },
    {
      "entropy": 5.734835052490235,
      "epoch": 2.478428321338261,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004380692604632424,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2070143774151802,
      "num_tokens": 55262226.0,
      "step": 31855
    },
    {
      "entropy": 5.780878400802612,
      "epoch": 2.4788173507099787,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043805016292178923,
      "loss": 5.0707,
      "mean_token_accuracy": 0.2065988853573799,
      "num_tokens": 55271523.0,
      "step": 31860
    },
    {
      "entropy": 5.748878860473633,
      "epoch": 2.4792063800816964,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004380310629063094,
      "loss": 4.9923,
      "mean_token_accuracy": 0.20676274448633195,
      "num_tokens": 55280604.0,
      "step": 31865
    },
    {
      "entropy": 5.627570915222168,
      "epoch": 2.4795954094534136,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004380119604170925,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2112988278269768,
      "num_tokens": 55289521.0,
      "step": 31870
    },
    {
      "entropy": 5.705765914916992,
      "epoch": 2.4799844388251313,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043799285545442876,
      "loss": 4.971,
      "mean_token_accuracy": 0.20409024208784105,
      "num_tokens": 55298211.0,
      "step": 31875
    },
    {
      "entropy": 5.678426074981689,
      "epoch": 2.480373468196849,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004379737480186078,
      "loss": 4.891,
      "mean_token_accuracy": 0.21074733436107634,
      "num_tokens": 55306984.0,
      "step": 31880
    },
    {
      "entropy": 5.730706214904785,
      "epoch": 2.4807624975685663,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043795463810991956,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20009558498859406,
      "num_tokens": 55316266.0,
      "step": 31885
    },
    {
      "entropy": 5.704307317733765,
      "epoch": 2.481151526940284,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004379355257286541,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21096517890691757,
      "num_tokens": 55324698.0,
      "step": 31890
    },
    {
      "entropy": 5.747517681121826,
      "epoch": 2.4815405563120017,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043791641087510136,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21254400312900543,
      "num_tokens": 55333033.0,
      "step": 31895
    },
    {
      "entropy": 5.757124662399292,
      "epoch": 2.4819295856837194,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004378972935495514,
      "loss": 5.0768,
      "mean_token_accuracy": 0.20114060938358308,
      "num_tokens": 55341110.0,
      "step": 31900
    },
    {
      "entropy": 5.748851728439331,
      "epoch": 2.4823186150554366,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004378781737522943,
      "loss": 4.9632,
      "mean_token_accuracy": 0.2036869779229164,
      "num_tokens": 55350156.0,
      "step": 31905
    },
    {
      "entropy": 5.748511791229248,
      "epoch": 2.4827076444271543,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043785905148362007,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20535746663808824,
      "num_tokens": 55358737.0,
      "step": 31910
    },
    {
      "entropy": 5.72822790145874,
      "epoch": 2.483096673798872,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043783992674381903,
      "loss": 4.9187,
      "mean_token_accuracy": 0.213743956387043,
      "num_tokens": 55366872.0,
      "step": 31915
    },
    {
      "entropy": 5.675046730041504,
      "epoch": 2.4834857031705893,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043782079953318126,
      "loss": 4.935,
      "mean_token_accuracy": 0.20745985805988312,
      "num_tokens": 55375675.0,
      "step": 31920
    },
    {
      "entropy": 5.731701469421386,
      "epoch": 2.483874732542307,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004378016698519971,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2087663695216179,
      "num_tokens": 55382750.0,
      "step": 31925
    },
    {
      "entropy": 5.6757824420928955,
      "epoch": 2.4842637619140246,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043778253770055666,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20920515358448027,
      "num_tokens": 55390852.0,
      "step": 31930
    },
    {
      "entropy": 5.6909082412719725,
      "epoch": 2.484652791285742,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043776340307915033,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20804184526205063,
      "num_tokens": 55398789.0,
      "step": 31935
    },
    {
      "entropy": 5.724300146102905,
      "epoch": 2.4850418206574596,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004377442659880684,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20294227600097656,
      "num_tokens": 55408100.0,
      "step": 31940
    },
    {
      "entropy": 5.727480792999268,
      "epoch": 2.4854308500291773,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004377251264276013,
      "loss": 4.959,
      "mean_token_accuracy": 0.21509731858968734,
      "num_tokens": 55416718.0,
      "step": 31945
    },
    {
      "entropy": 5.736598062515259,
      "epoch": 2.4858198794008945,
      "grad_norm": 1.25,
      "learning_rate": 0.00043770598439803936,
      "loss": 5.0435,
      "mean_token_accuracy": 0.2058554321527481,
      "num_tokens": 55426149.0,
      "step": 31950
    },
    {
      "entropy": 5.705715084075928,
      "epoch": 2.4862089087726122,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043768683989967304,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20415453612804413,
      "num_tokens": 55434379.0,
      "step": 31955
    },
    {
      "entropy": 5.701308965682983,
      "epoch": 2.48659793814433,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043766769293279294,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20421833395957947,
      "num_tokens": 55442880.0,
      "step": 31960
    },
    {
      "entropy": 5.756893062591553,
      "epoch": 2.4869869675160476,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043764854349768957,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2050406366586685,
      "num_tokens": 55452090.0,
      "step": 31965
    },
    {
      "entropy": 5.70805139541626,
      "epoch": 2.487375996887765,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043762939159465333,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2048500195145607,
      "num_tokens": 55460271.0,
      "step": 31970
    },
    {
      "entropy": 5.637308073043823,
      "epoch": 2.4877650262594826,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043761023722397495,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2123533546924591,
      "num_tokens": 55468414.0,
      "step": 31975
    },
    {
      "entropy": 5.720237731933594,
      "epoch": 2.4881540556312003,
      "grad_norm": 1.171875,
      "learning_rate": 0.000437591080385945,
      "loss": 4.9938,
      "mean_token_accuracy": 0.20741284042596816,
      "num_tokens": 55477658.0,
      "step": 31980
    },
    {
      "entropy": 5.7277015209197994,
      "epoch": 2.4885430850029175,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004375719210808543,
      "loss": 4.9195,
      "mean_token_accuracy": 0.20775629281997682,
      "num_tokens": 55486536.0,
      "step": 31985
    },
    {
      "entropy": 5.680407428741455,
      "epoch": 2.4889321143746352,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004375527593089934,
      "loss": 4.9865,
      "mean_token_accuracy": 0.20605216026306153,
      "num_tokens": 55494995.0,
      "step": 31990
    },
    {
      "entropy": 5.718147230148316,
      "epoch": 2.489321143746353,
      "grad_norm": 1.25,
      "learning_rate": 0.0004375335950706531,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19442166090011598,
      "num_tokens": 55503392.0,
      "step": 31995
    },
    {
      "entropy": 5.717027807235718,
      "epoch": 2.4897101731180706,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004375144283661242,
      "loss": 4.9632,
      "mean_token_accuracy": 0.21038024574518205,
      "num_tokens": 55512211.0,
      "step": 32000
    },
    {
      "entropy": 5.677935552597046,
      "epoch": 2.490099202489788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004374952591956975,
      "loss": 4.9411,
      "mean_token_accuracy": 0.20882178694009781,
      "num_tokens": 55520412.0,
      "step": 32005
    },
    {
      "entropy": 5.629581546783447,
      "epoch": 2.4904882318615056,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043747608755966396,
      "loss": 4.9394,
      "mean_token_accuracy": 0.20917407125234605,
      "num_tokens": 55529964.0,
      "step": 32010
    },
    {
      "entropy": 5.699930620193482,
      "epoch": 2.4908772612332233,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004374569134583143,
      "loss": 4.9585,
      "mean_token_accuracy": 0.20151448547840117,
      "num_tokens": 55539111.0,
      "step": 32015
    },
    {
      "entropy": 5.725243949890137,
      "epoch": 2.4912662906049405,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004374377368919396,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2147189974784851,
      "num_tokens": 55547520.0,
      "step": 32020
    },
    {
      "entropy": 5.744519186019898,
      "epoch": 2.4916553199766582,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043741855786083085,
      "loss": 5.006,
      "mean_token_accuracy": 0.2038813754916191,
      "num_tokens": 55557111.0,
      "step": 32025
    },
    {
      "entropy": 5.71839017868042,
      "epoch": 2.492044349348376,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043739937636527894,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20276594907045364,
      "num_tokens": 55565383.0,
      "step": 32030
    },
    {
      "entropy": 5.7687163829803465,
      "epoch": 2.492433378720093,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043738019240557494,
      "loss": 4.9864,
      "mean_token_accuracy": 0.2045573800802231,
      "num_tokens": 55573668.0,
      "step": 32035
    },
    {
      "entropy": 5.735308122634888,
      "epoch": 2.492822408091811,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043736100598201,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2051223635673523,
      "num_tokens": 55582716.0,
      "step": 32040
    },
    {
      "entropy": 5.6722198009490965,
      "epoch": 2.4932114374635286,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004373418170948752,
      "loss": 4.9488,
      "mean_token_accuracy": 0.2072313830256462,
      "num_tokens": 55591472.0,
      "step": 32045
    },
    {
      "entropy": 5.785842943191528,
      "epoch": 2.493600466835246,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004373226257444617,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21092396527528762,
      "num_tokens": 55600022.0,
      "step": 32050
    },
    {
      "entropy": 5.80219202041626,
      "epoch": 2.4939894962069635,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004373034319310608,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20357997268438338,
      "num_tokens": 55608854.0,
      "step": 32055
    },
    {
      "entropy": 5.627676773071289,
      "epoch": 2.4943785255786812,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043728423565496367,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20742380023002624,
      "num_tokens": 55618246.0,
      "step": 32060
    },
    {
      "entropy": 5.76334056854248,
      "epoch": 2.494767554950399,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004372650369164615,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20024004131555556,
      "num_tokens": 55626706.0,
      "step": 32065
    },
    {
      "entropy": 5.712987422943115,
      "epoch": 2.495156584322116,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043724583571584565,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2072004944086075,
      "num_tokens": 55634810.0,
      "step": 32070
    },
    {
      "entropy": 5.747103643417359,
      "epoch": 2.495545613693834,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043722663205340753,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20199114680290223,
      "num_tokens": 55643801.0,
      "step": 32075
    },
    {
      "entropy": 5.690426158905029,
      "epoch": 2.4959346430655516,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004372074259294384,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21192134469747542,
      "num_tokens": 55652236.0,
      "step": 32080
    },
    {
      "entropy": 5.720027017593384,
      "epoch": 2.496323672437269,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004371882173442299,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2100144773721695,
      "num_tokens": 55660996.0,
      "step": 32085
    },
    {
      "entropy": 5.751222705841064,
      "epoch": 2.4967127018089865,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043716900629807323,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19820431470870972,
      "num_tokens": 55669766.0,
      "step": 32090
    },
    {
      "entropy": 5.71356201171875,
      "epoch": 2.497101731180704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043714979279126006,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2076013684272766,
      "num_tokens": 55679179.0,
      "step": 32095
    },
    {
      "entropy": 5.754448747634887,
      "epoch": 2.497490760552422,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004371305768240818,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2032962992787361,
      "num_tokens": 55687554.0,
      "step": 32100
    },
    {
      "entropy": 5.650505018234253,
      "epoch": 2.497879789924139,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004371113583968302,
      "loss": 4.8765,
      "mean_token_accuracy": 0.21406447291374206,
      "num_tokens": 55695537.0,
      "step": 32105
    },
    {
      "entropy": 5.7501589298248295,
      "epoch": 2.498268819295857,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004370921375097967,
      "loss": 5.0626,
      "mean_token_accuracy": 0.20581903159618378,
      "num_tokens": 55703432.0,
      "step": 32110
    },
    {
      "entropy": 5.710600662231445,
      "epoch": 2.4986578486675746,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043707291416327307,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2063142627477646,
      "num_tokens": 55712169.0,
      "step": 32115
    },
    {
      "entropy": 5.766243839263916,
      "epoch": 2.499046878039292,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004370536883575509,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20191788375377656,
      "num_tokens": 55720471.0,
      "step": 32120
    },
    {
      "entropy": 5.700145816802978,
      "epoch": 2.4994359074110095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004370344600929219,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2087366446852684,
      "num_tokens": 55729188.0,
      "step": 32125
    },
    {
      "entropy": 5.702551031112671,
      "epoch": 2.499824936782727,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004370152293696779,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2075962096452713,
      "num_tokens": 55738313.0,
      "step": 32130
    },
    {
      "entropy": 5.708211231231689,
      "epoch": 2.500213966154445,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004369959961881107,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20476262867450715,
      "num_tokens": 55746849.0,
      "step": 32135
    },
    {
      "entropy": 5.709503841400147,
      "epoch": 2.500602995526162,
      "grad_norm": 1.28125,
      "learning_rate": 0.000436976760548512,
      "loss": 5.0179,
      "mean_token_accuracy": 0.20252264738082887,
      "num_tokens": 55755151.0,
      "step": 32140
    },
    {
      "entropy": 5.658016538619995,
      "epoch": 2.50099202489788,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043695752245117393,
      "loss": 4.9059,
      "mean_token_accuracy": 0.21205879002809525,
      "num_tokens": 55763716.0,
      "step": 32145
    },
    {
      "entropy": 5.711342763900757,
      "epoch": 2.501381054269597,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004369382818963882,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20347398221492768,
      "num_tokens": 55772424.0,
      "step": 32150
    },
    {
      "entropy": 5.76393141746521,
      "epoch": 2.501770083641315,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004369190388844467,
      "loss": 5.0959,
      "mean_token_accuracy": 0.19972964227199555,
      "num_tokens": 55780084.0,
      "step": 32155
    },
    {
      "entropy": 5.712643575668335,
      "epoch": 2.5021591130130325,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043689979341564154,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21048187762498854,
      "num_tokens": 55788717.0,
      "step": 32160
    },
    {
      "entropy": 5.698560333251953,
      "epoch": 2.50254814238475,
      "grad_norm": 1.25,
      "learning_rate": 0.0004368805454902648,
      "loss": 4.9504,
      "mean_token_accuracy": 0.20658242255449294,
      "num_tokens": 55797166.0,
      "step": 32165
    },
    {
      "entropy": 5.767977905273438,
      "epoch": 2.5029371717564675,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043686129510860826,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20500310212373735,
      "num_tokens": 55805702.0,
      "step": 32170
    },
    {
      "entropy": 5.675443696975708,
      "epoch": 2.503326201128185,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043684204227096435,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21179161816835404,
      "num_tokens": 55814938.0,
      "step": 32175
    },
    {
      "entropy": 5.712873125076294,
      "epoch": 2.503715230499903,
      "grad_norm": 1.203125,
      "learning_rate": 0.000436822786977625,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2049968957901001,
      "num_tokens": 55823527.0,
      "step": 32180
    },
    {
      "entropy": 5.814451885223389,
      "epoch": 2.50410425987162,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043680352922888244,
      "loss": 5.0355,
      "mean_token_accuracy": 0.20490658581256865,
      "num_tokens": 55831924.0,
      "step": 32185
    },
    {
      "entropy": 5.7139441013336185,
      "epoch": 2.504493289243338,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004367842690250288,
      "loss": 4.9429,
      "mean_token_accuracy": 0.21234404742717744,
      "num_tokens": 55840799.0,
      "step": 32190
    },
    {
      "entropy": 5.646137666702271,
      "epoch": 2.5048823186150555,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004367650063663565,
      "loss": 4.8895,
      "mean_token_accuracy": 0.21799386143684388,
      "num_tokens": 55848950.0,
      "step": 32195
    },
    {
      "entropy": 5.666651582717895,
      "epoch": 2.505271347986773,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004367457412531576,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21608682870864868,
      "num_tokens": 55857140.0,
      "step": 32200
    },
    {
      "entropy": 5.680488348007202,
      "epoch": 2.5056603773584905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004367264736857245,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20380420684814454,
      "num_tokens": 55865790.0,
      "step": 32205
    },
    {
      "entropy": 5.695819520950318,
      "epoch": 2.506049406730208,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004367072036643497,
      "loss": 4.9219,
      "mean_token_accuracy": 0.20680708289146424,
      "num_tokens": 55873598.0,
      "step": 32210
    },
    {
      "entropy": 5.717639207839966,
      "epoch": 2.506438436101926,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043668793118932536,
      "loss": 5.0581,
      "mean_token_accuracy": 0.1989049956202507,
      "num_tokens": 55882172.0,
      "step": 32215
    },
    {
      "entropy": 5.641362857818604,
      "epoch": 2.506827465473643,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043666865626094405,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20626555234193802,
      "num_tokens": 55890701.0,
      "step": 32220
    },
    {
      "entropy": 5.752888917922974,
      "epoch": 2.507216494845361,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043664937887949825,
      "loss": 5.062,
      "mean_token_accuracy": 0.19440430402755737,
      "num_tokens": 55899470.0,
      "step": 32225
    },
    {
      "entropy": 5.72113151550293,
      "epoch": 2.5076055242170785,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004366300990452803,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20356197357177735,
      "num_tokens": 55907442.0,
      "step": 32230
    },
    {
      "entropy": 5.732328224182129,
      "epoch": 2.507994553588796,
      "grad_norm": 1.25,
      "learning_rate": 0.00043661081675858295,
      "loss": 4.9914,
      "mean_token_accuracy": 0.20378621220588683,
      "num_tokens": 55915761.0,
      "step": 32235
    },
    {
      "entropy": 5.637710046768189,
      "epoch": 2.5083835829605134,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043659153201969865,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20434807240962982,
      "num_tokens": 55924101.0,
      "step": 32240
    },
    {
      "entropy": 5.773267078399658,
      "epoch": 2.508772612332231,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004365722448289201,
      "loss": 5.0995,
      "mean_token_accuracy": 0.2000497967004776,
      "num_tokens": 55933472.0,
      "step": 32245
    },
    {
      "entropy": 5.732623147964477,
      "epoch": 2.5091616417039484,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004365529551865399,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20024742931127548,
      "num_tokens": 55941962.0,
      "step": 32250
    },
    {
      "entropy": 5.703354358673096,
      "epoch": 2.509550671075666,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043653366309285066,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2038203537464142,
      "num_tokens": 55951392.0,
      "step": 32255
    },
    {
      "entropy": 5.701999139785767,
      "epoch": 2.509939700447384,
      "grad_norm": 1.25,
      "learning_rate": 0.0004365143685481452,
      "loss": 4.9462,
      "mean_token_accuracy": 0.20887251049280167,
      "num_tokens": 55959995.0,
      "step": 32260
    },
    {
      "entropy": 5.694582605361939,
      "epoch": 2.5103287298191015,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004364950715527164,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2024376630783081,
      "num_tokens": 55967960.0,
      "step": 32265
    },
    {
      "entropy": 5.714631986618042,
      "epoch": 2.5107177591908187,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004364757721068568,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20238029211759567,
      "num_tokens": 55976620.0,
      "step": 32270
    },
    {
      "entropy": 5.667682743072509,
      "epoch": 2.5111067885625364,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043645647021085945,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20744826942682265,
      "num_tokens": 55985478.0,
      "step": 32275
    },
    {
      "entropy": 5.697075223922729,
      "epoch": 2.511495817934254,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043643716586501706,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20539512634277343,
      "num_tokens": 55994123.0,
      "step": 32280
    },
    {
      "entropy": 5.733211708068848,
      "epoch": 2.5118848473059714,
      "grad_norm": 1.25,
      "learning_rate": 0.00043641785906962276,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2056300550699234,
      "num_tokens": 56003088.0,
      "step": 32285
    },
    {
      "entropy": 5.645250749588013,
      "epoch": 2.512273876677689,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004363985498249693,
      "loss": 4.9213,
      "mean_token_accuracy": 0.20981871485710143,
      "num_tokens": 56011349.0,
      "step": 32290
    },
    {
      "entropy": 5.66168098449707,
      "epoch": 2.512662906049407,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043637923813134974,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2156715899705887,
      "num_tokens": 56020225.0,
      "step": 32295
    },
    {
      "entropy": 5.776936864852905,
      "epoch": 2.5130519354211245,
      "grad_norm": 1.421875,
      "learning_rate": 0.00043635992398905713,
      "loss": 5.0204,
      "mean_token_accuracy": 0.20606142431497573,
      "num_tokens": 56028729.0,
      "step": 32300
    },
    {
      "entropy": 5.711764669418335,
      "epoch": 2.5134409647928417,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004363406073983844,
      "loss": 4.979,
      "mean_token_accuracy": 0.2017035037279129,
      "num_tokens": 56037173.0,
      "step": 32305
    },
    {
      "entropy": 5.704709434509278,
      "epoch": 2.5138299941645594,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043632128835962485,
      "loss": 5.0389,
      "mean_token_accuracy": 0.2024555280804634,
      "num_tokens": 56046013.0,
      "step": 32310
    },
    {
      "entropy": 5.772328567504883,
      "epoch": 2.514219023536277,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043630196687307156,
      "loss": 5.0715,
      "mean_token_accuracy": 0.2008786290884018,
      "num_tokens": 56054942.0,
      "step": 32315
    },
    {
      "entropy": 5.830512523651123,
      "epoch": 2.5146080529079944,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004362826429390176,
      "loss": 5.0928,
      "mean_token_accuracy": 0.19342033714056014,
      "num_tokens": 56064033.0,
      "step": 32320
    },
    {
      "entropy": 5.731173324584961,
      "epoch": 2.514997082279712,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043626331655775624,
      "loss": 5.0163,
      "mean_token_accuracy": 0.2089637488126755,
      "num_tokens": 56072534.0,
      "step": 32325
    },
    {
      "entropy": 5.695740127563477,
      "epoch": 2.51538611165143,
      "grad_norm": 1.25,
      "learning_rate": 0.00043624398772958075,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2102305546402931,
      "num_tokens": 56081742.0,
      "step": 32330
    },
    {
      "entropy": 5.701024341583252,
      "epoch": 2.5157751410231475,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004362246564547844,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21613372564315797,
      "num_tokens": 56090659.0,
      "step": 32335
    },
    {
      "entropy": 5.690423965454102,
      "epoch": 2.5161641703948647,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043620532273366044,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2037885919213295,
      "num_tokens": 56099795.0,
      "step": 32340
    },
    {
      "entropy": 5.670936822891235,
      "epoch": 2.5165531997665824,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004361859865665024,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20328075736761092,
      "num_tokens": 56107265.0,
      "step": 32345
    },
    {
      "entropy": 5.710996389389038,
      "epoch": 2.5169422291382997,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043616664795360346,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2108874350786209,
      "num_tokens": 56116434.0,
      "step": 32350
    },
    {
      "entropy": 5.751490879058838,
      "epoch": 2.5173312585100174,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043614730689525716,
      "loss": 5.0746,
      "mean_token_accuracy": 0.19641307443380357,
      "num_tokens": 56125356.0,
      "step": 32355
    },
    {
      "entropy": 5.787407112121582,
      "epoch": 2.517720287881735,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043612796339175707,
      "loss": 5.1026,
      "mean_token_accuracy": 0.20240970104932784,
      "num_tokens": 56133732.0,
      "step": 32360
    },
    {
      "entropy": 5.7639257431030275,
      "epoch": 2.5181093172534528,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043610861744339653,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21222862005233764,
      "num_tokens": 56142037.0,
      "step": 32365
    },
    {
      "entropy": 5.757498836517334,
      "epoch": 2.5184983466251705,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004360892690504692,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20342593789100646,
      "num_tokens": 56151064.0,
      "step": 32370
    },
    {
      "entropy": 5.722907304763794,
      "epoch": 2.5188873759968877,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004360699182132685,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2129116728901863,
      "num_tokens": 56159617.0,
      "step": 32375
    },
    {
      "entropy": 5.7119691371917725,
      "epoch": 2.5192764053686054,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043605056493208825,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20812912732362748,
      "num_tokens": 56168941.0,
      "step": 32380
    },
    {
      "entropy": 5.716279792785644,
      "epoch": 2.5196654347403227,
      "grad_norm": 1.21875,
      "learning_rate": 0.000436031209207222,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20544006526470185,
      "num_tokens": 56177321.0,
      "step": 32385
    },
    {
      "entropy": 5.721947097778321,
      "epoch": 2.5200544641120404,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043601185103896346,
      "loss": 5.027,
      "mean_token_accuracy": 0.19760787934064866,
      "num_tokens": 56185775.0,
      "step": 32390
    },
    {
      "entropy": 5.689810371398925,
      "epoch": 2.520443493483758,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043599249042760636,
      "loss": 5.0268,
      "mean_token_accuracy": 0.20442779809236528,
      "num_tokens": 56194949.0,
      "step": 32395
    },
    {
      "entropy": 5.7262975692749025,
      "epoch": 2.5208325228554758,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043597312737344446,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2026412308216095,
      "num_tokens": 56203314.0,
      "step": 32400
    },
    {
      "entropy": 5.712629556655884,
      "epoch": 2.521221552227193,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043595376187677157,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20730996131896973,
      "num_tokens": 56211857.0,
      "step": 32405
    },
    {
      "entropy": 5.691487836837768,
      "epoch": 2.5216105815989107,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043593439393788147,
      "loss": 4.9887,
      "mean_token_accuracy": 0.20606181621551514,
      "num_tokens": 56220331.0,
      "step": 32410
    },
    {
      "entropy": 5.727001571655274,
      "epoch": 2.5219996109706284,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004359150235570682,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20716377049684526,
      "num_tokens": 56228623.0,
      "step": 32415
    },
    {
      "entropy": 5.649047899246216,
      "epoch": 2.5223886403423457,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004358956507346255,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21276525259017945,
      "num_tokens": 56237028.0,
      "step": 32420
    },
    {
      "entropy": 5.65262999534607,
      "epoch": 2.5227776697140634,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004358762754708474,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20281088948249817,
      "num_tokens": 56244955.0,
      "step": 32425
    },
    {
      "entropy": 5.633661985397339,
      "epoch": 2.523166699085781,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043585689776602795,
      "loss": 4.968,
      "mean_token_accuracy": 0.20961283445358275,
      "num_tokens": 56254020.0,
      "step": 32430
    },
    {
      "entropy": 5.680561971664429,
      "epoch": 2.5235557284574988,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043583751762046104,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20646125227212905,
      "num_tokens": 56262829.0,
      "step": 32435
    },
    {
      "entropy": 5.7693342685699465,
      "epoch": 2.523944757829216,
      "grad_norm": 1.25,
      "learning_rate": 0.0004358181350344408,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2092021584510803,
      "num_tokens": 56271289.0,
      "step": 32440
    },
    {
      "entropy": 5.674266242980957,
      "epoch": 2.5243337872009337,
      "grad_norm": 1.25,
      "learning_rate": 0.0004357987500082613,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21365648061037062,
      "num_tokens": 56279701.0,
      "step": 32445
    },
    {
      "entropy": 5.73402042388916,
      "epoch": 2.524722816572651,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043577936254221675,
      "loss": 4.951,
      "mean_token_accuracy": 0.21067662984132768,
      "num_tokens": 56288373.0,
      "step": 32450
    },
    {
      "entropy": 5.70132851600647,
      "epoch": 2.5251118459443687,
      "grad_norm": 1.125,
      "learning_rate": 0.0004357599726366013,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20501175969839097,
      "num_tokens": 56297853.0,
      "step": 32455
    },
    {
      "entropy": 5.779371500015259,
      "epoch": 2.5255008753160864,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004357405802917091,
      "loss": 5.1255,
      "mean_token_accuracy": 0.19826317429542542,
      "num_tokens": 56306456.0,
      "step": 32460
    },
    {
      "entropy": 5.719353199005127,
      "epoch": 2.525889904687804,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004357211855078345,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2157145544886589,
      "num_tokens": 56314420.0,
      "step": 32465
    },
    {
      "entropy": 5.706261825561524,
      "epoch": 2.5262789340595218,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043570178828527157,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20157341361045839,
      "num_tokens": 56324011.0,
      "step": 32470
    },
    {
      "entropy": 5.6554368019104,
      "epoch": 2.526667963431239,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004356823886243149,
      "loss": 4.9311,
      "mean_token_accuracy": 0.20953617691993714,
      "num_tokens": 56332830.0,
      "step": 32475
    },
    {
      "entropy": 5.737408304214478,
      "epoch": 2.5270569928029567,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043566298652525875,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2125445693731308,
      "num_tokens": 56341486.0,
      "step": 32480
    },
    {
      "entropy": 5.750052261352539,
      "epoch": 2.527446022174674,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043564358198839746,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20673557817935945,
      "num_tokens": 56349382.0,
      "step": 32485
    },
    {
      "entropy": 5.712090682983399,
      "epoch": 2.5278350515463917,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004356241750140256,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20112408101558685,
      "num_tokens": 56358335.0,
      "step": 32490
    },
    {
      "entropy": 5.7501543998718265,
      "epoch": 2.5282240809181094,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043560476560243743,
      "loss": 5.0375,
      "mean_token_accuracy": 0.2046893820166588,
      "num_tokens": 56366777.0,
      "step": 32495
    },
    {
      "entropy": 5.799756383895874,
      "epoch": 2.528613110289827,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004355853537539276,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20503442436456681,
      "num_tokens": 56375175.0,
      "step": 32500
    },
    {
      "entropy": 5.709357643127442,
      "epoch": 2.5290021396615443,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004355659394687906,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2061050459742546,
      "num_tokens": 56384442.0,
      "step": 32505
    },
    {
      "entropy": 5.776636838912964,
      "epoch": 2.529391169033262,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004355465227473211,
      "loss": 5.0323,
      "mean_token_accuracy": 0.19949096143245698,
      "num_tokens": 56392420.0,
      "step": 32510
    },
    {
      "entropy": 5.749125862121582,
      "epoch": 2.5297801984049797,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004355271035898136,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20234971940517427,
      "num_tokens": 56400633.0,
      "step": 32515
    },
    {
      "entropy": 5.6973659038543705,
      "epoch": 2.530169227776697,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004355076819965628,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20949163883924485,
      "num_tokens": 56409606.0,
      "step": 32520
    },
    {
      "entropy": 5.7349710941314695,
      "epoch": 2.5305582571484146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043548825796786353,
      "loss": 5.0211,
      "mean_token_accuracy": 0.20994582027196884,
      "num_tokens": 56417724.0,
      "step": 32525
    },
    {
      "entropy": 5.705391454696655,
      "epoch": 2.5309472865201323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043546883150401023,
      "loss": 4.9856,
      "mean_token_accuracy": 0.2064082592725754,
      "num_tokens": 56426632.0,
      "step": 32530
    },
    {
      "entropy": 5.657213878631592,
      "epoch": 2.53133631589185,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043544940260529787,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2086200997233391,
      "num_tokens": 56435266.0,
      "step": 32535
    },
    {
      "entropy": 5.7342942237854,
      "epoch": 2.5317253452635673,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004354299712720213,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20974645912647247,
      "num_tokens": 56443335.0,
      "step": 32540
    },
    {
      "entropy": 5.7176989078521725,
      "epoch": 2.532114374635285,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004354105375044752,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2055739000439644,
      "num_tokens": 56452657.0,
      "step": 32545
    },
    {
      "entropy": 5.760034847259521,
      "epoch": 2.5325034040070027,
      "grad_norm": 1.25,
      "learning_rate": 0.00043539110130295445,
      "loss": 5.0489,
      "mean_token_accuracy": 0.19995319098234177,
      "num_tokens": 56461925.0,
      "step": 32550
    },
    {
      "entropy": 5.777818107604981,
      "epoch": 2.53289243337872,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043537166266775405,
      "loss": 4.9849,
      "mean_token_accuracy": 0.20698566734790802,
      "num_tokens": 56470356.0,
      "step": 32555
    },
    {
      "entropy": 5.69796724319458,
      "epoch": 2.5332814627504376,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000435352221599169,
      "loss": 5.0488,
      "mean_token_accuracy": 0.19879368990659713,
      "num_tokens": 56478754.0,
      "step": 32560
    },
    {
      "entropy": 5.6687414169311525,
      "epoch": 2.5336704921221553,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004353327780974941,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20892689973115922,
      "num_tokens": 56487089.0,
      "step": 32565
    },
    {
      "entropy": 5.752163314819336,
      "epoch": 2.534059521493873,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004353133321630246,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20068435817956926,
      "num_tokens": 56496416.0,
      "step": 32570
    },
    {
      "entropy": 5.690636730194091,
      "epoch": 2.5344485508655903,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004352938837960554,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2084551990032196,
      "num_tokens": 56505675.0,
      "step": 32575
    },
    {
      "entropy": 5.759299945831299,
      "epoch": 2.534837580237308,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043527443299688166,
      "loss": 4.9938,
      "mean_token_accuracy": 0.20754604637622834,
      "num_tokens": 56514855.0,
      "step": 32580
    },
    {
      "entropy": 5.812771606445312,
      "epoch": 2.5352266096090252,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043525497976579843,
      "loss": 5.067,
      "mean_token_accuracy": 0.20189114511013032,
      "num_tokens": 56524073.0,
      "step": 32585
    },
    {
      "entropy": 5.755522584915161,
      "epoch": 2.535615638980743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000435235524103101,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2105891987681389,
      "num_tokens": 56534127.0,
      "step": 32590
    },
    {
      "entropy": 5.86353554725647,
      "epoch": 2.5360046683524606,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004352160660090846,
      "loss": 5.102,
      "mean_token_accuracy": 0.19485899806022644,
      "num_tokens": 56543219.0,
      "step": 32595
    },
    {
      "entropy": 5.769903135299683,
      "epoch": 2.5363936977241783,
      "grad_norm": 1.25,
      "learning_rate": 0.0004351966054840443,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20924025326967238,
      "num_tokens": 56552146.0,
      "step": 32600
    },
    {
      "entropy": 5.7852250099182125,
      "epoch": 2.5367827270958956,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043517714252827556,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20097915977239608,
      "num_tokens": 56561780.0,
      "step": 32605
    },
    {
      "entropy": 5.682745122909546,
      "epoch": 2.5371717564676133,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004351576771420737,
      "loss": 4.9321,
      "mean_token_accuracy": 0.21296055763959884,
      "num_tokens": 56569864.0,
      "step": 32610
    },
    {
      "entropy": 5.8051185131073,
      "epoch": 2.537560785839331,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004351382093257339,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19691717624664307,
      "num_tokens": 56578506.0,
      "step": 32615
    },
    {
      "entropy": 5.773283815383911,
      "epoch": 2.5379498152110482,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004351187390795517,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21358393430709838,
      "num_tokens": 56587264.0,
      "step": 32620
    },
    {
      "entropy": 5.712436962127685,
      "epoch": 2.538338844582766,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043509926640382244,
      "loss": 4.8929,
      "mean_token_accuracy": 0.21315336525440215,
      "num_tokens": 56595779.0,
      "step": 32625
    },
    {
      "entropy": 5.6375892162323,
      "epoch": 2.5387278739544836,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004350797912988417,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21009058505296707,
      "num_tokens": 56604251.0,
      "step": 32630
    },
    {
      "entropy": 5.665219449996949,
      "epoch": 2.5391169033262013,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004350603137649049,
      "loss": 4.9657,
      "mean_token_accuracy": 0.20006470680236815,
      "num_tokens": 56612527.0,
      "step": 32635
    },
    {
      "entropy": 5.753927993774414,
      "epoch": 2.5395059326979186,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004350408338023075,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20302853286266326,
      "num_tokens": 56621139.0,
      "step": 32640
    },
    {
      "entropy": 5.73028564453125,
      "epoch": 2.5398949620696363,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004350213514113453,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21223442554473876,
      "num_tokens": 56629757.0,
      "step": 32645
    },
    {
      "entropy": 5.646002340316772,
      "epoch": 2.540283991441354,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043500186659231377,
      "loss": 4.8945,
      "mean_token_accuracy": 0.20956345796585082,
      "num_tokens": 56638518.0,
      "step": 32650
    },
    {
      "entropy": 5.681959581375122,
      "epoch": 2.5406730208130712,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004349823793455086,
      "loss": 4.9943,
      "mean_token_accuracy": 0.19662700444459916,
      "num_tokens": 56647542.0,
      "step": 32655
    },
    {
      "entropy": 5.7725427627563475,
      "epoch": 2.541062050184789,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043496288967122547,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2034281998872757,
      "num_tokens": 56655771.0,
      "step": 32660
    },
    {
      "entropy": 5.682890319824219,
      "epoch": 2.5414510795565066,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043494339756976007,
      "loss": 4.9182,
      "mean_token_accuracy": 0.20901809334754945,
      "num_tokens": 56663895.0,
      "step": 32665
    },
    {
      "entropy": 5.642807149887085,
      "epoch": 2.5418401089282243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004349239030414082,
      "loss": 4.991,
      "mean_token_accuracy": 0.20504894554615022,
      "num_tokens": 56672557.0,
      "step": 32670
    },
    {
      "entropy": 5.7123250484466555,
      "epoch": 2.5422291382999416,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043490440608646557,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2004629299044609,
      "num_tokens": 56681095.0,
      "step": 32675
    },
    {
      "entropy": 5.779498052597046,
      "epoch": 2.5426181676716593,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043488490670522817,
      "loss": 5.063,
      "mean_token_accuracy": 0.20217254012823105,
      "num_tokens": 56689347.0,
      "step": 32680
    },
    {
      "entropy": 5.680990934371948,
      "epoch": 2.5430071970433765,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004348654048979918,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21077237278223038,
      "num_tokens": 56698348.0,
      "step": 32685
    },
    {
      "entropy": 5.709836053848266,
      "epoch": 2.543396226415094,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004348459006650523,
      "loss": 5.0397,
      "mean_token_accuracy": 0.20332515984773636,
      "num_tokens": 56707215.0,
      "step": 32690
    },
    {
      "entropy": 5.728739309310913,
      "epoch": 2.543785255786812,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043482639400670576,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20101978182792662,
      "num_tokens": 56716190.0,
      "step": 32695
    },
    {
      "entropy": 5.691095542907715,
      "epoch": 2.5441742851585296,
      "grad_norm": 1.265625,
      "learning_rate": 0.000434806884923248,
      "loss": 4.9339,
      "mean_token_accuracy": 0.20973406434059144,
      "num_tokens": 56724494.0,
      "step": 32700
    },
    {
      "entropy": 5.749963617324829,
      "epoch": 2.5445633145302473,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004347873734149752,
      "loss": 4.9396,
      "mean_token_accuracy": 0.21079889982938765,
      "num_tokens": 56732648.0,
      "step": 32705
    },
    {
      "entropy": 5.824687433242798,
      "epoch": 2.5449523439019646,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043476785948218325,
      "loss": 5.1438,
      "mean_token_accuracy": 0.1967386409640312,
      "num_tokens": 56742270.0,
      "step": 32710
    },
    {
      "entropy": 5.741421699523926,
      "epoch": 2.5453413732736823,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043474834312516835,
      "loss": 4.988,
      "mean_token_accuracy": 0.20593938380479812,
      "num_tokens": 56750982.0,
      "step": 32715
    },
    {
      "entropy": 5.680449962615967,
      "epoch": 2.5457304026453995,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004347288243442266,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21222972720861435,
      "num_tokens": 56759443.0,
      "step": 32720
    },
    {
      "entropy": 5.758137273788452,
      "epoch": 2.546119432017117,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004347093031396543,
      "loss": 5.0161,
      "mean_token_accuracy": 0.20138444155454635,
      "num_tokens": 56768704.0,
      "step": 32725
    },
    {
      "entropy": 5.72704930305481,
      "epoch": 2.546508461388835,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004346897795117474,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2099384054541588,
      "num_tokens": 56777274.0,
      "step": 32730
    },
    {
      "entropy": 5.756817817687988,
      "epoch": 2.5468974907605526,
      "grad_norm": 1.25,
      "learning_rate": 0.0004346702534608023,
      "loss": 5.0074,
      "mean_token_accuracy": 0.19940920919179916,
      "num_tokens": 56786184.0,
      "step": 32735
    },
    {
      "entropy": 5.720722198486328,
      "epoch": 2.54728652013227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004346507249871153,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2050049126148224,
      "num_tokens": 56795156.0,
      "step": 32740
    },
    {
      "entropy": 5.776710033416748,
      "epoch": 2.5476755495039876,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004346311940909826,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20134371668100357,
      "num_tokens": 56803829.0,
      "step": 32745
    },
    {
      "entropy": 5.721598672866821,
      "epoch": 2.5480645788757053,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043461166077270056,
      "loss": 4.9896,
      "mean_token_accuracy": 0.21170148104429246,
      "num_tokens": 56813191.0,
      "step": 32750
    },
    {
      "entropy": 5.769086599349976,
      "epoch": 2.5484536082474225,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043459212503256566,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20273256599903106,
      "num_tokens": 56822338.0,
      "step": 32755
    },
    {
      "entropy": 5.764503765106201,
      "epoch": 2.54884263761914,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004345725868708743,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20435421615839006,
      "num_tokens": 56830873.0,
      "step": 32760
    },
    {
      "entropy": 5.722545051574707,
      "epoch": 2.549231666990858,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043455304628792294,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2045636370778084,
      "num_tokens": 56839462.0,
      "step": 32765
    },
    {
      "entropy": 5.755924367904663,
      "epoch": 2.5496206963625756,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000434533503284008,
      "loss": 5.0167,
      "mean_token_accuracy": 0.20389361381530763,
      "num_tokens": 56847713.0,
      "step": 32770
    },
    {
      "entropy": 5.780830144882202,
      "epoch": 2.550009725734293,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004345139578594261,
      "loss": 4.9401,
      "mean_token_accuracy": 0.20662273913621904,
      "num_tokens": 56856507.0,
      "step": 32775
    },
    {
      "entropy": 5.76229076385498,
      "epoch": 2.5503987551060106,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004344944100144738,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21500105857849122,
      "num_tokens": 56864772.0,
      "step": 32780
    },
    {
      "entropy": 5.763923406600952,
      "epoch": 2.550787784477728,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043447485974944765,
      "loss": 5.0249,
      "mean_token_accuracy": 0.1981954127550125,
      "num_tokens": 56872886.0,
      "step": 32785
    },
    {
      "entropy": 5.740024423599243,
      "epoch": 2.5511768138494455,
      "grad_norm": 1.25,
      "learning_rate": 0.00043445530706464437,
      "loss": 5.0026,
      "mean_token_accuracy": 0.20340923964977264,
      "num_tokens": 56882079.0,
      "step": 32790
    },
    {
      "entropy": 5.675230836868286,
      "epoch": 2.551565843221163,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004344357519603606,
      "loss": 4.9416,
      "mean_token_accuracy": 0.21266695857048035,
      "num_tokens": 56890237.0,
      "step": 32795
    },
    {
      "entropy": 5.743769311904908,
      "epoch": 2.551954872592881,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004344161944368931,
      "loss": 5.0327,
      "mean_token_accuracy": 0.1989602416753769,
      "num_tokens": 56898703.0,
      "step": 32800
    },
    {
      "entropy": 5.802595329284668,
      "epoch": 2.5523439019645986,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004343966344945385,
      "loss": 5.1086,
      "mean_token_accuracy": 0.20099483877420427,
      "num_tokens": 56907954.0,
      "step": 32805
    },
    {
      "entropy": 5.621290063858032,
      "epoch": 2.552732931336316,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043437707213359375,
      "loss": 4.8552,
      "mean_token_accuracy": 0.21395466923713685,
      "num_tokens": 56916316.0,
      "step": 32810
    },
    {
      "entropy": 5.654263210296631,
      "epoch": 2.5531219607080335,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043435750735435554,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21547530591487885,
      "num_tokens": 56924386.0,
      "step": 32815
    },
    {
      "entropy": 5.786268901824951,
      "epoch": 2.553510990079751,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043433794015712085,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21374816596508026,
      "num_tokens": 56932782.0,
      "step": 32820
    },
    {
      "entropy": 5.742829132080078,
      "epoch": 2.5539000194514685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004343183705421865,
      "loss": 4.945,
      "mean_token_accuracy": 0.19924988448619843,
      "num_tokens": 56941030.0,
      "step": 32825
    },
    {
      "entropy": 5.836594200134277,
      "epoch": 2.554289048823186,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004342987985098495,
      "loss": 5.1848,
      "mean_token_accuracy": 0.1947267085313797,
      "num_tokens": 56949525.0,
      "step": 32830
    },
    {
      "entropy": 5.668847465515137,
      "epoch": 2.554678078194904,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043427922406040676,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20840259045362472,
      "num_tokens": 56958317.0,
      "step": 32835
    },
    {
      "entropy": 5.71341028213501,
      "epoch": 2.555067107566621,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043425964719415525,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20657522976398468,
      "num_tokens": 56966848.0,
      "step": 32840
    },
    {
      "entropy": 5.785091781616211,
      "epoch": 2.555456136938339,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043424006791139213,
      "loss": 5.0659,
      "mean_token_accuracy": 0.2017503947019577,
      "num_tokens": 56975448.0,
      "step": 32845
    },
    {
      "entropy": 5.809026765823364,
      "epoch": 2.5558451663100565,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043422048621241443,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20007745027542115,
      "num_tokens": 56983800.0,
      "step": 32850
    },
    {
      "entropy": 5.781334161758423,
      "epoch": 2.556234195681774,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043420090209751933,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21078215539455414,
      "num_tokens": 56992505.0,
      "step": 32855
    },
    {
      "entropy": 5.680140733718872,
      "epoch": 2.5566232250534915,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004341813155670039,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21927272230386735,
      "num_tokens": 57000471.0,
      "step": 32860
    },
    {
      "entropy": 5.768496036529541,
      "epoch": 2.557012254425209,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004341617266211654,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20630973428487778,
      "num_tokens": 57008708.0,
      "step": 32865
    },
    {
      "entropy": 5.824241209030151,
      "epoch": 2.557401283796927,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000434142135260301,
      "loss": 5.0944,
      "mean_token_accuracy": 0.2017227977514267,
      "num_tokens": 57017395.0,
      "step": 32870
    },
    {
      "entropy": 5.83169732093811,
      "epoch": 2.557790313168644,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000434122541484708,
      "loss": 4.9623,
      "mean_token_accuracy": 0.1998027518391609,
      "num_tokens": 57025535.0,
      "step": 32875
    },
    {
      "entropy": 5.714866065979004,
      "epoch": 2.558179342540362,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004341029452946837,
      "loss": 4.9432,
      "mean_token_accuracy": 0.20742350667715073,
      "num_tokens": 57034029.0,
      "step": 32880
    },
    {
      "entropy": 5.6944386005401615,
      "epoch": 2.558568371912079,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004340833466905254,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20113589614629745,
      "num_tokens": 57042779.0,
      "step": 32885
    },
    {
      "entropy": 5.786851406097412,
      "epoch": 2.558957401283797,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043406374567253055,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2046095222234726,
      "num_tokens": 57052506.0,
      "step": 32890
    },
    {
      "entropy": 5.794452428817749,
      "epoch": 2.5593464306555145,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004340441422409965,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19647629112005233,
      "num_tokens": 57061307.0,
      "step": 32895
    },
    {
      "entropy": 5.7341409683227536,
      "epoch": 2.559735460027232,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004340245363962209,
      "loss": 5.0931,
      "mean_token_accuracy": 0.20352228879928588,
      "num_tokens": 57069991.0,
      "step": 32900
    },
    {
      "entropy": 5.717057991027832,
      "epoch": 2.56012448939895,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043400492813850083,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2068238824605942,
      "num_tokens": 57077768.0,
      "step": 32905
    },
    {
      "entropy": 5.700071239471436,
      "epoch": 2.560513518770667,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043398531746813426,
      "loss": 4.9624,
      "mean_token_accuracy": 0.20671363323926925,
      "num_tokens": 57086666.0,
      "step": 32910
    },
    {
      "entropy": 5.705666494369507,
      "epoch": 2.560902548142385,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043396570438541845,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20829079449176788,
      "num_tokens": 57095352.0,
      "step": 32915
    },
    {
      "entropy": 5.672802829742432,
      "epoch": 2.561291577514102,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004339460888906512,
      "loss": 4.9515,
      "mean_token_accuracy": 0.2129378840327263,
      "num_tokens": 57104535.0,
      "step": 32920
    },
    {
      "entropy": 5.778772830963135,
      "epoch": 2.5616806068858198,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004339264709841299,
      "loss": 4.9808,
      "mean_token_accuracy": 0.202763369679451,
      "num_tokens": 57112523.0,
      "step": 32925
    },
    {
      "entropy": 5.651504373550415,
      "epoch": 2.5620696362575375,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043390685066615244,
      "loss": 4.8977,
      "mean_token_accuracy": 0.2071441352367401,
      "num_tokens": 57120921.0,
      "step": 32930
    },
    {
      "entropy": 5.716360092163086,
      "epoch": 2.562458665629255,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004338872279370164,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20127545893192292,
      "num_tokens": 57128398.0,
      "step": 32935
    },
    {
      "entropy": 5.72646050453186,
      "epoch": 2.5628476950009724,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004338676027970196,
      "loss": 5.0196,
      "mean_token_accuracy": 0.20369046032428742,
      "num_tokens": 57136888.0,
      "step": 32940
    },
    {
      "entropy": 5.724941873550415,
      "epoch": 2.56323672437269,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043384797524645977,
      "loss": 4.9733,
      "mean_token_accuracy": 0.20686656534671782,
      "num_tokens": 57144795.0,
      "step": 32945
    },
    {
      "entropy": 5.772236633300781,
      "epoch": 2.563625753744408,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004338283452856348,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19889288693666457,
      "num_tokens": 57153073.0,
      "step": 32950
    },
    {
      "entropy": 5.7662616729736325,
      "epoch": 2.564014783116125,
      "grad_norm": 1.25,
      "learning_rate": 0.0004338087129148425,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20340803861618043,
      "num_tokens": 57161536.0,
      "step": 32955
    },
    {
      "entropy": 5.780566453933716,
      "epoch": 2.5644038124878428,
      "grad_norm": 1.375,
      "learning_rate": 0.00043378907813438066,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20063025504350662,
      "num_tokens": 57170065.0,
      "step": 32960
    },
    {
      "entropy": 5.6704993724823,
      "epoch": 2.5647928418595605,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00043376944094454734,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2131212681531906,
      "num_tokens": 57179063.0,
      "step": 32965
    },
    {
      "entropy": 5.647791385650635,
      "epoch": 2.565181871231278,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043374980134564044,
      "loss": 4.8364,
      "mean_token_accuracy": 0.22899574786424637,
      "num_tokens": 57187803.0,
      "step": 32970
    },
    {
      "entropy": 5.771262073516846,
      "epoch": 2.5655709006029954,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043373015933795804,
      "loss": 5.1001,
      "mean_token_accuracy": 0.19359012246131896,
      "num_tokens": 57196428.0,
      "step": 32975
    },
    {
      "entropy": 5.770413446426391,
      "epoch": 2.565959929974713,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004337105149217981,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20524235665798188,
      "num_tokens": 57205541.0,
      "step": 32980
    },
    {
      "entropy": 5.709564256668091,
      "epoch": 2.566348959346431,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043369086809745875,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21073568612337112,
      "num_tokens": 57214436.0,
      "step": 32985
    },
    {
      "entropy": 5.730032920837402,
      "epoch": 2.566737988718148,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043367121886523795,
      "loss": 5.0103,
      "mean_token_accuracy": 0.19986904859542848,
      "num_tokens": 57222359.0,
      "step": 32990
    },
    {
      "entropy": 5.663343477249145,
      "epoch": 2.5671270180898658,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000433651567225434,
      "loss": 4.9089,
      "mean_token_accuracy": 0.20501043051481246,
      "num_tokens": 57230226.0,
      "step": 32995
    },
    {
      "entropy": 5.6929929733276365,
      "epoch": 2.5675160474615835,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004336319131783451,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20198056548833848,
      "num_tokens": 57239585.0,
      "step": 33000
    },
    {
      "epoch": 2.5675160474615835,
      "eval_entropy": 5.441885370744663,
      "eval_loss": 5.078895092010498,
      "eval_mean_token_accuracy": 0.21144095331975424,
      "eval_num_tokens": 57239585.0,
      "eval_runtime": 21.7987,
      "eval_samples_per_second": 1906.447,
      "eval_steps_per_second": 238.317,
      "step": 33000
    },
    {
      "entropy": 5.690830945968628,
      "epoch": 2.567905076833301,
      "grad_norm": 1.25,
      "learning_rate": 0.00043361225672426933,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20790625214576722,
      "num_tokens": 57248108.0,
      "step": 33005
    },
    {
      "entropy": 5.780340909957886,
      "epoch": 2.5682941062050184,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004335925978635051,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20567653477191924,
      "num_tokens": 57257758.0,
      "step": 33010
    },
    {
      "entropy": 5.794309186935425,
      "epoch": 2.568683135576736,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043357293659635057,
      "loss": 5.0411,
      "mean_token_accuracy": 0.20353172570466996,
      "num_tokens": 57267180.0,
      "step": 33015
    },
    {
      "entropy": 5.708866357803345,
      "epoch": 2.5690721649484534,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004335532729231041,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20778950452804565,
      "num_tokens": 57275756.0,
      "step": 33020
    },
    {
      "entropy": 5.743973684310913,
      "epoch": 2.569461194320171,
      "grad_norm": 1.375,
      "learning_rate": 0.00043353360684406415,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2027055263519287,
      "num_tokens": 57284593.0,
      "step": 33025
    },
    {
      "entropy": 5.7824639797210695,
      "epoch": 2.5698502236918888,
      "grad_norm": 1.25,
      "learning_rate": 0.000433513938359529,
      "loss": 5.047,
      "mean_token_accuracy": 0.20221317708492278,
      "num_tokens": 57293304.0,
      "step": 33030
    },
    {
      "entropy": 5.701857423782348,
      "epoch": 2.5702392530636065,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004334942674697971,
      "loss": 4.992,
      "mean_token_accuracy": 0.21032871305942535,
      "num_tokens": 57302057.0,
      "step": 33035
    },
    {
      "entropy": 5.738317823410034,
      "epoch": 2.5706282824353237,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043347459417516696,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20367086827754974,
      "num_tokens": 57310576.0,
      "step": 33040
    },
    {
      "entropy": 5.773430871963501,
      "epoch": 2.5710173118070414,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004334549184759371,
      "loss": 4.9917,
      "mean_token_accuracy": 0.20841752588748932,
      "num_tokens": 57319103.0,
      "step": 33045
    },
    {
      "entropy": 5.832637357711792,
      "epoch": 2.571406341178759,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004334352403724062,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20312875509262085,
      "num_tokens": 57329075.0,
      "step": 33050
    },
    {
      "entropy": 5.791800212860108,
      "epoch": 2.5717953705504764,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043341555986487254,
      "loss": 4.947,
      "mean_token_accuracy": 0.2069276601076126,
      "num_tokens": 57337986.0,
      "step": 33055
    },
    {
      "entropy": 5.751558637619018,
      "epoch": 2.572184399922194,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004333958769536349,
      "loss": 5.0804,
      "mean_token_accuracy": 0.19780918806791306,
      "num_tokens": 57346440.0,
      "step": 33060
    },
    {
      "entropy": 5.746549415588379,
      "epoch": 2.5725734292939118,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004333761916389921,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19772604405879973,
      "num_tokens": 57354791.0,
      "step": 33065
    },
    {
      "entropy": 5.7414055347442625,
      "epoch": 2.5729624586656294,
      "grad_norm": 1.25,
      "learning_rate": 0.0004333565039212425,
      "loss": 4.9855,
      "mean_token_accuracy": 0.20605745017528534,
      "num_tokens": 57363200.0,
      "step": 33070
    },
    {
      "entropy": 5.714569902420044,
      "epoch": 2.5733514880373467,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004333368138006851,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2068865105509758,
      "num_tokens": 57371830.0,
      "step": 33075
    },
    {
      "entropy": 5.806778812408448,
      "epoch": 2.5737405174090644,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004333171212776185,
      "loss": 5.0894,
      "mean_token_accuracy": 0.19771674573421477,
      "num_tokens": 57380123.0,
      "step": 33080
    },
    {
      "entropy": 5.790821075439453,
      "epoch": 2.574129546780782,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004332974263523416,
      "loss": 5.0904,
      "mean_token_accuracy": 0.19944725334644317,
      "num_tokens": 57389093.0,
      "step": 33085
    },
    {
      "entropy": 5.615280961990356,
      "epoch": 2.5745185761524993,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043327772902515327,
      "loss": 4.8159,
      "mean_token_accuracy": 0.22546641975641252,
      "num_tokens": 57397165.0,
      "step": 33090
    },
    {
      "entropy": 5.7218705177307125,
      "epoch": 2.574907605524217,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004332580292963523,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19963347166776657,
      "num_tokens": 57406308.0,
      "step": 33095
    },
    {
      "entropy": 5.724903726577759,
      "epoch": 2.5752966348959347,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004332383271662376,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21200001388788223,
      "num_tokens": 57414964.0,
      "step": 33100
    },
    {
      "entropy": 5.63138370513916,
      "epoch": 2.5756856642676524,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043321862263510816,
      "loss": 4.8357,
      "mean_token_accuracy": 0.22915955781936645,
      "num_tokens": 57422684.0,
      "step": 33105
    },
    {
      "entropy": 5.652907371520996,
      "epoch": 2.5760746936393697,
      "grad_norm": 1.125,
      "learning_rate": 0.0004331989157032629,
      "loss": 4.9509,
      "mean_token_accuracy": 0.20198525488376617,
      "num_tokens": 57431784.0,
      "step": 33110
    },
    {
      "entropy": 5.723780775070191,
      "epoch": 2.5764637230110874,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043317920637100097,
      "loss": 5.0462,
      "mean_token_accuracy": 0.1981985479593277,
      "num_tokens": 57440467.0,
      "step": 33115
    },
    {
      "entropy": 5.728650856018066,
      "epoch": 2.5768527523828046,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004331594946386213,
      "loss": 4.9392,
      "mean_token_accuracy": 0.20674220770597457,
      "num_tokens": 57449263.0,
      "step": 33120
    },
    {
      "entropy": 5.783505153656006,
      "epoch": 2.5772417817545223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043313978050642303,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21379772424697877,
      "num_tokens": 57457662.0,
      "step": 33125
    },
    {
      "entropy": 5.715204477310181,
      "epoch": 2.57763081112624,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004331200639747053,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20954393446445466,
      "num_tokens": 57466049.0,
      "step": 33130
    },
    {
      "entropy": 5.708579063415527,
      "epoch": 2.5780198404979577,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004331003450437674,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20443544536828995,
      "num_tokens": 57474923.0,
      "step": 33135
    },
    {
      "entropy": 5.704727268218994,
      "epoch": 2.5784088698696754,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004330806237139083,
      "loss": 4.9335,
      "mean_token_accuracy": 0.20291880071163176,
      "num_tokens": 57483292.0,
      "step": 33140
    },
    {
      "entropy": 5.697311544418335,
      "epoch": 2.5787978992413927,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004330608999854273,
      "loss": 4.9832,
      "mean_token_accuracy": 0.20623408108949662,
      "num_tokens": 57491894.0,
      "step": 33145
    },
    {
      "entropy": 5.7016129970550535,
      "epoch": 2.5791869286131104,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004330411738586238,
      "loss": 5.0158,
      "mean_token_accuracy": 0.20100595206022262,
      "num_tokens": 57500999.0,
      "step": 33150
    },
    {
      "entropy": 5.819173574447632,
      "epoch": 2.5795759579848276,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000433021445333797,
      "loss": 5.0952,
      "mean_token_accuracy": 0.1956484317779541,
      "num_tokens": 57508964.0,
      "step": 33155
    },
    {
      "entropy": 5.810405111312866,
      "epoch": 2.5799649873565453,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043300171441124633,
      "loss": 5.0845,
      "mean_token_accuracy": 0.20089659094810486,
      "num_tokens": 57517605.0,
      "step": 33160
    },
    {
      "entropy": 5.838309335708618,
      "epoch": 2.580354016728263,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004329819810912711,
      "loss": 5.1607,
      "mean_token_accuracy": 0.1942279666662216,
      "num_tokens": 57527043.0,
      "step": 33165
    },
    {
      "entropy": 5.787107467651367,
      "epoch": 2.5807430460999807,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043296224537417075,
      "loss": 4.9978,
      "mean_token_accuracy": 0.2077171564102173,
      "num_tokens": 57535768.0,
      "step": 33170
    },
    {
      "entropy": 5.765929555892944,
      "epoch": 2.581132075471698,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043294250726024473,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2006270855665207,
      "num_tokens": 57544060.0,
      "step": 33175
    },
    {
      "entropy": 5.690450143814087,
      "epoch": 2.5815211048434157,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004329227667497926,
      "loss": 4.8953,
      "mean_token_accuracy": 0.21223667562007903,
      "num_tokens": 57553139.0,
      "step": 33180
    },
    {
      "entropy": 5.703609132766724,
      "epoch": 2.5819101342151334,
      "grad_norm": 1.25,
      "learning_rate": 0.00043290302384311373,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2050335392355919,
      "num_tokens": 57561388.0,
      "step": 33185
    },
    {
      "entropy": 5.697948455810547,
      "epoch": 2.5822991635868506,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043288327854050794,
      "loss": 4.958,
      "mean_token_accuracy": 0.20468744039535522,
      "num_tokens": 57570431.0,
      "step": 33190
    },
    {
      "entropy": 5.75230450630188,
      "epoch": 2.5826881929585683,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043286353084227467,
      "loss": 4.9992,
      "mean_token_accuracy": 0.19813869446516036,
      "num_tokens": 57579047.0,
      "step": 33195
    },
    {
      "entropy": 5.82316746711731,
      "epoch": 2.583077222330286,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043284378074871354,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20098772048950195,
      "num_tokens": 57587878.0,
      "step": 33200
    },
    {
      "entropy": 5.772847270965576,
      "epoch": 2.5834662517020037,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004328240282601243,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21217236816883087,
      "num_tokens": 57596823.0,
      "step": 33205
    },
    {
      "entropy": 5.758020639419556,
      "epoch": 2.583855281073721,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004328042733768066,
      "loss": 5.1106,
      "mean_token_accuracy": 0.19461209028959275,
      "num_tokens": 57605983.0,
      "step": 33210
    },
    {
      "entropy": 5.749850940704346,
      "epoch": 2.5842443104454387,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043278451609906027,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20652249902486802,
      "num_tokens": 57613987.0,
      "step": 33215
    },
    {
      "entropy": 5.768683767318725,
      "epoch": 2.584633339817156,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043276475642718503,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20777551531791688,
      "num_tokens": 57622409.0,
      "step": 33220
    },
    {
      "entropy": 5.74588532447815,
      "epoch": 2.5850223691888736,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004327449943614808,
      "loss": 4.986,
      "mean_token_accuracy": 0.20727429687976837,
      "num_tokens": 57630664.0,
      "step": 33225
    },
    {
      "entropy": 5.7309496879577635,
      "epoch": 2.5854113985605913,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043272522990224727,
      "loss": 5.0277,
      "mean_token_accuracy": 0.19899268299341202,
      "num_tokens": 57639745.0,
      "step": 33230
    },
    {
      "entropy": 5.689175796508789,
      "epoch": 2.585800427932309,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004327054630497844,
      "loss": 4.9767,
      "mean_token_accuracy": 0.21248547583818436,
      "num_tokens": 57648628.0,
      "step": 33235
    },
    {
      "entropy": 5.766215467453003,
      "epoch": 2.5861894573040267,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043268569380439223,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2223304897546768,
      "num_tokens": 57656322.0,
      "step": 33240
    },
    {
      "entropy": 5.747747230529785,
      "epoch": 2.586578486675744,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004326659221663705,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2063898652791977,
      "num_tokens": 57664807.0,
      "step": 33245
    },
    {
      "entropy": 5.675253105163574,
      "epoch": 2.5869675160474617,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004326461481360194,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20590822994709015,
      "num_tokens": 57673216.0,
      "step": 33250
    },
    {
      "entropy": 5.723646879196167,
      "epoch": 2.587356545419179,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043262637171363906,
      "loss": 5.0564,
      "mean_token_accuracy": 0.20054297000169755,
      "num_tokens": 57681615.0,
      "step": 33255
    },
    {
      "entropy": 5.892910194396973,
      "epoch": 2.5877455747908966,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043260659289952926,
      "loss": 5.0891,
      "mean_token_accuracy": 0.2085789129137993,
      "num_tokens": 57690951.0,
      "step": 33260
    },
    {
      "entropy": 5.766534805297852,
      "epoch": 2.5881346041626143,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043258681169399034,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2040807530283928,
      "num_tokens": 57699723.0,
      "step": 33265
    },
    {
      "entropy": 5.717220163345337,
      "epoch": 2.588523633534332,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043256702809732234,
      "loss": 4.9757,
      "mean_token_accuracy": 0.2047529086470604,
      "num_tokens": 57708380.0,
      "step": 33270
    },
    {
      "entropy": 5.78683910369873,
      "epoch": 2.5889126629060493,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004325472421098255,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20539262890815735,
      "num_tokens": 57717183.0,
      "step": 33275
    },
    {
      "entropy": 5.746925926208496,
      "epoch": 2.589301692277767,
      "grad_norm": 1.125,
      "learning_rate": 0.00043252745373180006,
      "loss": 4.9508,
      "mean_token_accuracy": 0.20586524307727813,
      "num_tokens": 57726369.0,
      "step": 33280
    },
    {
      "entropy": 5.777370309829712,
      "epoch": 2.5896907216494847,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004325076629635462,
      "loss": 4.9541,
      "mean_token_accuracy": 0.21324413120746613,
      "num_tokens": 57734863.0,
      "step": 33285
    },
    {
      "entropy": 5.721614837646484,
      "epoch": 2.590079751021202,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043248786980536424,
      "loss": 4.9836,
      "mean_token_accuracy": 0.2151501327753067,
      "num_tokens": 57743478.0,
      "step": 33290
    },
    {
      "entropy": 5.721319675445557,
      "epoch": 2.5904687803929196,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004324680742575545,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20620955377817154,
      "num_tokens": 57752300.0,
      "step": 33295
    },
    {
      "entropy": 5.753612422943116,
      "epoch": 2.5908578097646373,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043244827632041744,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2032677173614502,
      "num_tokens": 57760377.0,
      "step": 33300
    },
    {
      "entropy": 5.854331064224243,
      "epoch": 2.591246839136355,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004324284759942534,
      "loss": 5.1791,
      "mean_token_accuracy": 0.18954287618398666,
      "num_tokens": 57768665.0,
      "step": 33305
    },
    {
      "entropy": 5.730392932891846,
      "epoch": 2.5916358685080723,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004324086732793627,
      "loss": 4.9907,
      "mean_token_accuracy": 0.20835434049367904,
      "num_tokens": 57777387.0,
      "step": 33310
    },
    {
      "entropy": 5.832428407669068,
      "epoch": 2.59202489787979,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043238886817604605,
      "loss": 5.0814,
      "mean_token_accuracy": 0.20075132548809052,
      "num_tokens": 57785406.0,
      "step": 33315
    },
    {
      "entropy": 5.727300691604614,
      "epoch": 2.5924139272515077,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004323690606846039,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20731637626886368,
      "num_tokens": 57794412.0,
      "step": 33320
    },
    {
      "entropy": 5.7290692806243895,
      "epoch": 2.592802956623225,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043234925080533657,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2072455480694771,
      "num_tokens": 57803310.0,
      "step": 33325
    },
    {
      "entropy": 5.713151931762695,
      "epoch": 2.5931919859949426,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004323294385385448,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20227113366127014,
      "num_tokens": 57812968.0,
      "step": 33330
    },
    {
      "entropy": 5.7846180438995365,
      "epoch": 2.5935810153666603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004323096238845293,
      "loss": 5.1395,
      "mean_token_accuracy": 0.19564296752214433,
      "num_tokens": 57821230.0,
      "step": 33335
    },
    {
      "entropy": 5.820528221130371,
      "epoch": 2.593970044738378,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004322898068435906,
      "loss": 5.0703,
      "mean_token_accuracy": 0.2033640667796135,
      "num_tokens": 57829643.0,
      "step": 33340
    },
    {
      "entropy": 5.741799545288086,
      "epoch": 2.5943590741100953,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004322699874160294,
      "loss": 4.9518,
      "mean_token_accuracy": 0.20625571310520172,
      "num_tokens": 57837994.0,
      "step": 33345
    },
    {
      "entropy": 5.761845397949219,
      "epoch": 2.594748103481813,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043225016560214654,
      "loss": 5.0482,
      "mean_token_accuracy": 0.19883679449558259,
      "num_tokens": 57846764.0,
      "step": 33350
    },
    {
      "entropy": 5.760121202468872,
      "epoch": 2.59513713285353,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043223034140224266,
      "loss": 4.956,
      "mean_token_accuracy": 0.20872276723384858,
      "num_tokens": 57854887.0,
      "step": 33355
    },
    {
      "entropy": 5.773990345001221,
      "epoch": 2.595526162225248,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004322105148166187,
      "loss": 4.9651,
      "mean_token_accuracy": 0.20492430180311202,
      "num_tokens": 57863639.0,
      "step": 33360
    },
    {
      "entropy": 5.711425733566284,
      "epoch": 2.5959151915969656,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043219068584557526,
      "loss": 4.9247,
      "mean_token_accuracy": 0.20476991385221482,
      "num_tokens": 57872037.0,
      "step": 33365
    },
    {
      "entropy": 5.782890748977661,
      "epoch": 2.5963042209686833,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004321708544894134,
      "loss": 5.0253,
      "mean_token_accuracy": 0.20535183995962142,
      "num_tokens": 57880809.0,
      "step": 33370
    },
    {
      "entropy": 5.811325883865356,
      "epoch": 2.5966932503404005,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043215102074843403,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20585858821868896,
      "num_tokens": 57889184.0,
      "step": 33375
    },
    {
      "entropy": 5.7251636505126955,
      "epoch": 2.5970822797121182,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043213118462293796,
      "loss": 4.9461,
      "mean_token_accuracy": 0.20560525506734847,
      "num_tokens": 57897695.0,
      "step": 33380
    },
    {
      "entropy": 5.792908477783203,
      "epoch": 2.597471309083836,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004321113461132264,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20509863942861556,
      "num_tokens": 57906280.0,
      "step": 33385
    },
    {
      "entropy": 5.78499436378479,
      "epoch": 2.597860338455553,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043209150521960016,
      "loss": 5.03,
      "mean_token_accuracy": 0.20490869432687758,
      "num_tokens": 57915112.0,
      "step": 33390
    },
    {
      "entropy": 5.7358497142791744,
      "epoch": 2.598249367827271,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043207166194236037,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20105579644441604,
      "num_tokens": 57924210.0,
      "step": 33395
    },
    {
      "entropy": 5.767873048782349,
      "epoch": 2.5986383971989886,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004320518162818082,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20905519723892213,
      "num_tokens": 57933030.0,
      "step": 33400
    },
    {
      "entropy": 5.738175725936889,
      "epoch": 2.5990274265707063,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004320319682382445,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2027512565255165,
      "num_tokens": 57941490.0,
      "step": 33405
    },
    {
      "entropy": 5.6997240543365475,
      "epoch": 2.5994164559424235,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043201211781197075,
      "loss": 4.982,
      "mean_token_accuracy": 0.2057149514555931,
      "num_tokens": 57950666.0,
      "step": 33410
    },
    {
      "entropy": 5.712323713302612,
      "epoch": 2.5998054853141412,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043199226500328805,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20615371763706208,
      "num_tokens": 57959740.0,
      "step": 33415
    },
    {
      "entropy": 5.745664596557617,
      "epoch": 2.600194514685859,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004319724098124975,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20296220630407333,
      "num_tokens": 57968391.0,
      "step": 33420
    },
    {
      "entropy": 5.759553623199463,
      "epoch": 2.600583544057576,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043195255223990055,
      "loss": 5.0312,
      "mean_token_accuracy": 0.2043713301420212,
      "num_tokens": 57976873.0,
      "step": 33425
    },
    {
      "entropy": 5.796790075302124,
      "epoch": 2.600972573429294,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004319326922857984,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20630945712327958,
      "num_tokens": 57985085.0,
      "step": 33430
    },
    {
      "entropy": 5.778571224212646,
      "epoch": 2.6013616028010116,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004319128299504925,
      "loss": 5.0018,
      "mean_token_accuracy": 0.20420226454734802,
      "num_tokens": 57993686.0,
      "step": 33435
    },
    {
      "entropy": 5.792569351196289,
      "epoch": 2.6017506321727293,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043189296523428407,
      "loss": 5.1085,
      "mean_token_accuracy": 0.19711407870054246,
      "num_tokens": 58001877.0,
      "step": 33440
    },
    {
      "entropy": 5.74989881515503,
      "epoch": 2.6021396615444465,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043187309813747464,
      "loss": 4.9773,
      "mean_token_accuracy": 0.21105746626853944,
      "num_tokens": 58010688.0,
      "step": 33445
    },
    {
      "entropy": 5.7975705623626705,
      "epoch": 2.6025286909161642,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043185322866036565,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21008139103651047,
      "num_tokens": 58019470.0,
      "step": 33450
    },
    {
      "entropy": 5.819345283508301,
      "epoch": 2.6029177202878815,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043183335680325866,
      "loss": 5.0671,
      "mean_token_accuracy": 0.19122105836868286,
      "num_tokens": 58028297.0,
      "step": 33455
    },
    {
      "entropy": 5.749031496047974,
      "epoch": 2.603306749659599,
      "grad_norm": 1.25,
      "learning_rate": 0.00043181348256645497,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19399728029966354,
      "num_tokens": 58036815.0,
      "step": 33460
    },
    {
      "entropy": 5.82499213218689,
      "epoch": 2.603695779031317,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043179360595025637,
      "loss": 5.0831,
      "mean_token_accuracy": 0.2040332227945328,
      "num_tokens": 58045682.0,
      "step": 33465
    },
    {
      "entropy": 5.7769935607910154,
      "epoch": 2.6040848084030346,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004317737269549643,
      "loss": 5.0485,
      "mean_token_accuracy": 0.199982351064682,
      "num_tokens": 58054530.0,
      "step": 33470
    },
    {
      "entropy": 5.79452953338623,
      "epoch": 2.604473837774752,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004317538455808805,
      "loss": 5.067,
      "mean_token_accuracy": 0.20311780422925949,
      "num_tokens": 58063360.0,
      "step": 33475
    },
    {
      "entropy": 5.783742046356201,
      "epoch": 2.6048628671464695,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004317339618283065,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20600496083498002,
      "num_tokens": 58071198.0,
      "step": 33480
    },
    {
      "entropy": 5.824437284469605,
      "epoch": 2.6052518965181872,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004317140756975442,
      "loss": 5.0827,
      "mean_token_accuracy": 0.20267838090658188,
      "num_tokens": 58080138.0,
      "step": 33485
    },
    {
      "entropy": 5.729674673080444,
      "epoch": 2.6056409258899045,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004316941871888951,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21093521863222123,
      "num_tokens": 58088768.0,
      "step": 33490
    },
    {
      "entropy": 5.681346845626831,
      "epoch": 2.606029955261622,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043167429630266134,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2090834468603134,
      "num_tokens": 58098309.0,
      "step": 33495
    },
    {
      "entropy": 5.7474932193756105,
      "epoch": 2.60641898463334,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004316544030391444,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20776279866695405,
      "num_tokens": 58106987.0,
      "step": 33500
    },
    {
      "entropy": 5.764124059677124,
      "epoch": 2.6068080140050576,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004316345073986461,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2078765168786049,
      "num_tokens": 58115065.0,
      "step": 33505
    },
    {
      "entropy": 5.76682481765747,
      "epoch": 2.607197043376775,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004316146093814686,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20227448940277098,
      "num_tokens": 58123995.0,
      "step": 33510
    },
    {
      "entropy": 5.713196277618408,
      "epoch": 2.6075860727484925,
      "grad_norm": 1.25,
      "learning_rate": 0.0004315947089879137,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21625824719667436,
      "num_tokens": 58132080.0,
      "step": 33515
    },
    {
      "entropy": 5.772599983215332,
      "epoch": 2.60797510212021,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043157480621828327,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2049530878663063,
      "num_tokens": 58140199.0,
      "step": 33520
    },
    {
      "entropy": 5.665989732742309,
      "epoch": 2.6083641314919275,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043155490107287935,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21832228302955628,
      "num_tokens": 58148852.0,
      "step": 33525
    },
    {
      "entropy": 5.740784072875977,
      "epoch": 2.608753160863645,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004315349935520041,
      "loss": 5.0565,
      "mean_token_accuracy": 0.21023191511631012,
      "num_tokens": 58157290.0,
      "step": 33530
    },
    {
      "entropy": 5.780831527709961,
      "epoch": 2.609142190235363,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004315150836559594,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20357503443956376,
      "num_tokens": 58166086.0,
      "step": 33535
    },
    {
      "entropy": 5.781947612762451,
      "epoch": 2.6095312196070806,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004314951713850474,
      "loss": 4.9827,
      "mean_token_accuracy": 0.20581308752298355,
      "num_tokens": 58174865.0,
      "step": 33540
    },
    {
      "entropy": 5.72998538017273,
      "epoch": 2.609920248978798,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043147525673957026,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20490460097789764,
      "num_tokens": 58182313.0,
      "step": 33545
    },
    {
      "entropy": 5.753555154800415,
      "epoch": 2.6103092783505155,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043145533971983025,
      "loss": 5.0312,
      "mean_token_accuracy": 0.19593823105096816,
      "num_tokens": 58190758.0,
      "step": 33550
    },
    {
      "entropy": 5.791784477233887,
      "epoch": 2.6106983077222328,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043143542032612935,
      "loss": 5.031,
      "mean_token_accuracy": 0.20053247809410096,
      "num_tokens": 58199308.0,
      "step": 33555
    },
    {
      "entropy": 5.745243644714355,
      "epoch": 2.6110873370939505,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004314154985587701,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2010275349020958,
      "num_tokens": 58206995.0,
      "step": 33560
    },
    {
      "entropy": 5.694269847869873,
      "epoch": 2.611476366465668,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043139557441805457,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21183932721614837,
      "num_tokens": 58215186.0,
      "step": 33565
    },
    {
      "entropy": 5.761064910888672,
      "epoch": 2.611865395837386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043137564790428503,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20171750634908675,
      "num_tokens": 58224347.0,
      "step": 33570
    },
    {
      "entropy": 5.693391513824463,
      "epoch": 2.6122544252091036,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043135571901776397,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21420842707157134,
      "num_tokens": 58232477.0,
      "step": 33575
    },
    {
      "entropy": 5.6766197204589846,
      "epoch": 2.612643454580821,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004313357877587937,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21675673872232437,
      "num_tokens": 58241186.0,
      "step": 33580
    },
    {
      "entropy": 5.736015653610229,
      "epoch": 2.6130324839525385,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004313158541276768,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20904262065887452,
      "num_tokens": 58249140.0,
      "step": 33585
    },
    {
      "entropy": 5.735386371612549,
      "epoch": 2.6134215133242558,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004312959181247156,
      "loss": 4.9898,
      "mean_token_accuracy": 0.203854937851429,
      "num_tokens": 58258313.0,
      "step": 33590
    },
    {
      "entropy": 5.782800817489624,
      "epoch": 2.6138105426959735,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004312759797502126,
      "loss": 4.9784,
      "mean_token_accuracy": 0.20691556334495545,
      "num_tokens": 58267013.0,
      "step": 33595
    },
    {
      "entropy": 5.759813928604126,
      "epoch": 2.614199572067691,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043125603900447026,
      "loss": 5.0427,
      "mean_token_accuracy": 0.2035905033349991,
      "num_tokens": 58275844.0,
      "step": 33600
    },
    {
      "entropy": 5.767997169494629,
      "epoch": 2.614588601439409,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004312360958877913,
      "loss": 5.0377,
      "mean_token_accuracy": 0.19740207493305206,
      "num_tokens": 58284416.0,
      "step": 33605
    },
    {
      "entropy": 5.762695074081421,
      "epoch": 2.614977630811126,
      "grad_norm": 1.25,
      "learning_rate": 0.0004312161504004782,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20518822073936463,
      "num_tokens": 58293564.0,
      "step": 33610
    },
    {
      "entropy": 5.7148042678833,
      "epoch": 2.615366660182844,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043119620254283375,
      "loss": 4.9742,
      "mean_token_accuracy": 0.20095243006944657,
      "num_tokens": 58302230.0,
      "step": 33615
    },
    {
      "entropy": 5.776022434234619,
      "epoch": 2.6157556895545615,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043117625231516054,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20458514392375945,
      "num_tokens": 58309540.0,
      "step": 33620
    },
    {
      "entropy": 5.693380689620971,
      "epoch": 2.6161447189262788,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004311562997177612,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21420613080263137,
      "num_tokens": 58317885.0,
      "step": 33625
    },
    {
      "entropy": 5.754724168777466,
      "epoch": 2.6165337482979965,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004311363447509386,
      "loss": 5.0121,
      "mean_token_accuracy": 0.19465384632349014,
      "num_tokens": 58326133.0,
      "step": 33630
    },
    {
      "entropy": 5.750039196014404,
      "epoch": 2.616922777669714,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004311163874149955,
      "loss": 5.0061,
      "mean_token_accuracy": 0.2041172832250595,
      "num_tokens": 58335085.0,
      "step": 33635
    },
    {
      "entropy": 5.773416376113891,
      "epoch": 2.617311807041432,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043109642771023464,
      "loss": 5.025,
      "mean_token_accuracy": 0.20453732162714006,
      "num_tokens": 58344222.0,
      "step": 33640
    },
    {
      "entropy": 5.712661790847778,
      "epoch": 2.617700836413149,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004310764656369591,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2108660891652107,
      "num_tokens": 58353256.0,
      "step": 33645
    },
    {
      "entropy": 5.808213424682617,
      "epoch": 2.618089865784867,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004310565011954715,
      "loss": 5.0251,
      "mean_token_accuracy": 0.2119701847434044,
      "num_tokens": 58361197.0,
      "step": 33650
    },
    {
      "entropy": 5.800841426849365,
      "epoch": 2.618478895156584,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004310365343860748,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20904432088136674,
      "num_tokens": 58370653.0,
      "step": 33655
    },
    {
      "entropy": 5.761476325988769,
      "epoch": 2.6188679245283017,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043101656520907225,
      "loss": 5.0633,
      "mean_token_accuracy": 0.20085494816303254,
      "num_tokens": 58379804.0,
      "step": 33660
    },
    {
      "entropy": 5.755903005599976,
      "epoch": 2.6192569539000194,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004309965936647665,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20302470326423644,
      "num_tokens": 58388443.0,
      "step": 33665
    },
    {
      "entropy": 5.7995068550109865,
      "epoch": 2.619645983271737,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004309766197534608,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20163162648677826,
      "num_tokens": 58397165.0,
      "step": 33670
    },
    {
      "entropy": 5.779026222229004,
      "epoch": 2.620035012643455,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043095664347545816,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20409030765295028,
      "num_tokens": 58405679.0,
      "step": 33675
    },
    {
      "entropy": 5.716444063186645,
      "epoch": 2.620424042015172,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004309366648310616,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2142328754067421,
      "num_tokens": 58414998.0,
      "step": 33680
    },
    {
      "entropy": 5.7853577613830565,
      "epoch": 2.62081307138689,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043091668382057443,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21133146286010743,
      "num_tokens": 58423535.0,
      "step": 33685
    },
    {
      "entropy": 5.708835029602051,
      "epoch": 2.621202100758607,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043089670044429973,
      "loss": 4.9435,
      "mean_token_accuracy": 0.20400387048721313,
      "num_tokens": 58432379.0,
      "step": 33690
    },
    {
      "entropy": 5.689995050430298,
      "epoch": 2.6215911301303247,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043087671470254076,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20030051916837693,
      "num_tokens": 58441013.0,
      "step": 33695
    },
    {
      "entropy": 5.720353269577027,
      "epoch": 2.6219801595020424,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043085672659560077,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20521938800811768,
      "num_tokens": 58450353.0,
      "step": 33700
    },
    {
      "entropy": 5.797730302810669,
      "epoch": 2.62236918887376,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043083673612378295,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2011021703481674,
      "num_tokens": 58458833.0,
      "step": 33705
    },
    {
      "entropy": 5.754546976089477,
      "epoch": 2.6227582182454774,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043081674328739066,
      "loss": 5.0656,
      "mean_token_accuracy": 0.1965584486722946,
      "num_tokens": 58468325.0,
      "step": 33710
    },
    {
      "entropy": 5.712673616409302,
      "epoch": 2.623147247617195,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043079674808672744,
      "loss": 4.9478,
      "mean_token_accuracy": 0.20859015733003616,
      "num_tokens": 58476333.0,
      "step": 33715
    },
    {
      "entropy": 5.686579465866089,
      "epoch": 2.623536276988913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004307767505220964,
      "loss": 5.0022,
      "mean_token_accuracy": 0.2089739754796028,
      "num_tokens": 58484435.0,
      "step": 33720
    },
    {
      "entropy": 5.716101455688476,
      "epoch": 2.62392530636063,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004307567505938012,
      "loss": 5.0885,
      "mean_token_accuracy": 0.19650800824165343,
      "num_tokens": 58493709.0,
      "step": 33725
    },
    {
      "entropy": 5.766401052474976,
      "epoch": 2.6243143357323477,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004307367483021452,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20818150639533997,
      "num_tokens": 58502522.0,
      "step": 33730
    },
    {
      "entropy": 5.735060024261474,
      "epoch": 2.6247033651040654,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000430716743647432,
      "loss": 5.0129,
      "mean_token_accuracy": 0.2017705351114273,
      "num_tokens": 58511459.0,
      "step": 33735
    },
    {
      "entropy": 5.75135440826416,
      "epoch": 2.625092394475783,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000430696736629965,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20370592027902604,
      "num_tokens": 58520197.0,
      "step": 33740
    },
    {
      "entropy": 5.748404121398925,
      "epoch": 2.6254814238475004,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004306767272500478,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20431021898984908,
      "num_tokens": 58528587.0,
      "step": 33745
    },
    {
      "entropy": 5.769442510604859,
      "epoch": 2.625870453219218,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043065671550798415,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2069698691368103,
      "num_tokens": 58537348.0,
      "step": 33750
    },
    {
      "entropy": 5.684375619888305,
      "epoch": 2.626259482590936,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043063670140407753,
      "loss": 4.9014,
      "mean_token_accuracy": 0.2180325448513031,
      "num_tokens": 58545637.0,
      "step": 33755
    },
    {
      "entropy": 5.718052864074707,
      "epoch": 2.626648511962653,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004306166849386317,
      "loss": 4.9984,
      "mean_token_accuracy": 0.20036183595657348,
      "num_tokens": 58554575.0,
      "step": 33760
    },
    {
      "entropy": 5.666042757034302,
      "epoch": 2.6270375413343707,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043059666611195037,
      "loss": 4.8534,
      "mean_token_accuracy": 0.21641640812158586,
      "num_tokens": 58563150.0,
      "step": 33765
    },
    {
      "entropy": 5.684116840362549,
      "epoch": 2.6274265707060884,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004305766449243372,
      "loss": 5.0084,
      "mean_token_accuracy": 0.19867660403251647,
      "num_tokens": 58571790.0,
      "step": 33770
    },
    {
      "entropy": 5.717001056671142,
      "epoch": 2.627815600077806,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004305566213760962,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2094368413090706,
      "num_tokens": 58581094.0,
      "step": 33775
    },
    {
      "entropy": 5.775350761413574,
      "epoch": 2.6282046294495234,
      "grad_norm": 1.25,
      "learning_rate": 0.00043053659546753094,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2037036657333374,
      "num_tokens": 58590836.0,
      "step": 33780
    },
    {
      "entropy": 5.747936534881592,
      "epoch": 2.628593658821241,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004305165671989455,
      "loss": 4.9965,
      "mean_token_accuracy": 0.20642196983098984,
      "num_tokens": 58599315.0,
      "step": 33785
    },
    {
      "entropy": 5.712173271179199,
      "epoch": 2.6289826881929583,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004304965365706437,
      "loss": 4.9707,
      "mean_token_accuracy": 0.2102762907743454,
      "num_tokens": 58607663.0,
      "step": 33790
    },
    {
      "entropy": 5.684770202636718,
      "epoch": 2.629371717564676,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004304765035829294,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20619403570890427,
      "num_tokens": 58616513.0,
      "step": 33795
    },
    {
      "entropy": 5.802895784378052,
      "epoch": 2.6297607469363937,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043045646823610664,
      "loss": 5.0855,
      "mean_token_accuracy": 0.20475125610828399,
      "num_tokens": 58625483.0,
      "step": 33800
    },
    {
      "entropy": 5.7778599739074705,
      "epoch": 2.6301497763081114,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043043643053047935,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20563886910676957,
      "num_tokens": 58633880.0,
      "step": 33805
    },
    {
      "entropy": 5.7896003246307375,
      "epoch": 2.6305388056798287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004304163904663517,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20515665858983995,
      "num_tokens": 58642940.0,
      "step": 33810
    },
    {
      "entropy": 5.68115005493164,
      "epoch": 2.6309278350515464,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043039634804402767,
      "loss": 4.9446,
      "mean_token_accuracy": 0.211849245429039,
      "num_tokens": 58651736.0,
      "step": 33815
    },
    {
      "entropy": 5.7687578201293945,
      "epoch": 2.631316864423264,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004303763032638114,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20961988419294358,
      "num_tokens": 58660494.0,
      "step": 33820
    },
    {
      "entropy": 5.710988473892212,
      "epoch": 2.6317058937949813,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004303562561260071,
      "loss": 4.969,
      "mean_token_accuracy": 0.20412998497486115,
      "num_tokens": 58669150.0,
      "step": 33825
    },
    {
      "entropy": 5.694502353668213,
      "epoch": 2.632094923166699,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043033620663091883,
      "loss": 4.9241,
      "mean_token_accuracy": 0.2171587884426117,
      "num_tokens": 58677806.0,
      "step": 33830
    },
    {
      "entropy": 5.692282009124756,
      "epoch": 2.6324839525384167,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004303161547788509,
      "loss": 4.9407,
      "mean_token_accuracy": 0.20978717356920243,
      "num_tokens": 58686406.0,
      "step": 33835
    },
    {
      "entropy": 5.798107528686524,
      "epoch": 2.6328729819101344,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004302961005701074,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20807095170021056,
      "num_tokens": 58696074.0,
      "step": 33840
    },
    {
      "entropy": 5.783406972885132,
      "epoch": 2.6332620112818517,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043027604400499295,
      "loss": 5.0182,
      "mean_token_accuracy": 0.20302368998527526,
      "num_tokens": 58704860.0,
      "step": 33845
    },
    {
      "entropy": 5.770510911941528,
      "epoch": 2.6336510406535694,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043025598508381155,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2053261548280716,
      "num_tokens": 58713455.0,
      "step": 33850
    },
    {
      "entropy": 5.641428279876709,
      "epoch": 2.634040070025287,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004302359238068677,
      "loss": 4.9193,
      "mean_token_accuracy": 0.20828338861465454,
      "num_tokens": 58722382.0,
      "step": 33855
    },
    {
      "entropy": 5.730730867385864,
      "epoch": 2.6344290993970043,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043021586017446585,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21251511722803115,
      "num_tokens": 58730816.0,
      "step": 33860
    },
    {
      "entropy": 5.733964490890503,
      "epoch": 2.634818128768722,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043019579418691027,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2093673899769783,
      "num_tokens": 58740143.0,
      "step": 33865
    },
    {
      "entropy": 5.652185010910034,
      "epoch": 2.6352071581404397,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004301757258445056,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21312558352947236,
      "num_tokens": 58748347.0,
      "step": 33870
    },
    {
      "entropy": 5.6538341522216795,
      "epoch": 2.6355961875121574,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004301556551475563,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20123799741268159,
      "num_tokens": 58756274.0,
      "step": 33875
    },
    {
      "entropy": 5.721617078781128,
      "epoch": 2.6359852168838747,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004301355820963669,
      "loss": 4.9628,
      "mean_token_accuracy": 0.20844414085149765,
      "num_tokens": 58764564.0,
      "step": 33880
    },
    {
      "entropy": 5.6531532287597654,
      "epoch": 2.6363742462555924,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004301155066912419,
      "loss": 5.0179,
      "mean_token_accuracy": 0.19796012938022614,
      "num_tokens": 58772914.0,
      "step": 33885
    },
    {
      "entropy": 5.7461851119995115,
      "epoch": 2.6367632756273096,
      "grad_norm": 1.296875,
      "learning_rate": 0.000430095428932486,
      "loss": 5.0065,
      "mean_token_accuracy": 0.2033153295516968,
      "num_tokens": 58781572.0,
      "step": 33890
    },
    {
      "entropy": 5.7461306095123295,
      "epoch": 2.6371523049990273,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043007534882040385,
      "loss": 4.982,
      "mean_token_accuracy": 0.21042746305465698,
      "num_tokens": 58789259.0,
      "step": 33895
    },
    {
      "entropy": 5.75700945854187,
      "epoch": 2.637541334370745,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004300552663553001,
      "loss": 5.0103,
      "mean_token_accuracy": 0.21726167500019072,
      "num_tokens": 58797984.0,
      "step": 33900
    },
    {
      "entropy": 5.729117965698242,
      "epoch": 2.6379303637424627,
      "grad_norm": 1.25,
      "learning_rate": 0.0004300351815374795,
      "loss": 4.906,
      "mean_token_accuracy": 0.2220124199986458,
      "num_tokens": 58806474.0,
      "step": 33905
    },
    {
      "entropy": 5.7382818222045895,
      "epoch": 2.63831939311418,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004300150943672467,
      "loss": 5.0625,
      "mean_token_accuracy": 0.19953990131616592,
      "num_tokens": 58816223.0,
      "step": 33910
    },
    {
      "entropy": 5.751334476470947,
      "epoch": 2.6387084224858977,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004299950048449067,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2008117839694023,
      "num_tokens": 58824950.0,
      "step": 33915
    },
    {
      "entropy": 5.750315189361572,
      "epoch": 2.6390974518576154,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004299749129707641,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20906038582324982,
      "num_tokens": 58833201.0,
      "step": 33920
    },
    {
      "entropy": 5.742064428329468,
      "epoch": 2.6394864812293326,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004299548187451239,
      "loss": 4.944,
      "mean_token_accuracy": 0.20985999405384065,
      "num_tokens": 58841621.0,
      "step": 33925
    },
    {
      "entropy": 5.708352375030517,
      "epoch": 2.6398755106010503,
      "grad_norm": 1.375,
      "learning_rate": 0.00042993472216829097,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2100432813167572,
      "num_tokens": 58849831.0,
      "step": 33930
    },
    {
      "entropy": 5.696486711502075,
      "epoch": 2.640264539972768,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042991462324057025,
      "loss": 5.0222,
      "mean_token_accuracy": 0.20560503154993057,
      "num_tokens": 58858530.0,
      "step": 33935
    },
    {
      "entropy": 5.728975820541382,
      "epoch": 2.6406535693444857,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004298945219622667,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19541842639446258,
      "num_tokens": 58868127.0,
      "step": 33940
    },
    {
      "entropy": 5.783413076400757,
      "epoch": 2.641042598716203,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042987441833368525,
      "loss": 4.9585,
      "mean_token_accuracy": 0.20653257220983506,
      "num_tokens": 58876959.0,
      "step": 33945
    },
    {
      "entropy": 5.726172304153442,
      "epoch": 2.6414316280879206,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042985431235513104,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21058230698108674,
      "num_tokens": 58885534.0,
      "step": 33950
    },
    {
      "entropy": 5.767559909820557,
      "epoch": 2.6418206574596383,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042983420402690917,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20818566232919694,
      "num_tokens": 58894600.0,
      "step": 33955
    },
    {
      "entropy": 5.756769895553589,
      "epoch": 2.6422096868313556,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042981409334932457,
      "loss": 5.0218,
      "mean_token_accuracy": 0.2084101051092148,
      "num_tokens": 58903243.0,
      "step": 33960
    },
    {
      "entropy": 5.778383588790893,
      "epoch": 2.6425987162030733,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004297939803226826,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20628530979156495,
      "num_tokens": 58911058.0,
      "step": 33965
    },
    {
      "entropy": 5.745448160171509,
      "epoch": 2.642987745574791,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004297738649472884,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20710175335407258,
      "num_tokens": 58919698.0,
      "step": 33970
    },
    {
      "entropy": 5.856704378128052,
      "epoch": 2.6433767749465087,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042975374722344713,
      "loss": 5.1511,
      "mean_token_accuracy": 0.19114880412817,
      "num_tokens": 58929276.0,
      "step": 33975
    },
    {
      "entropy": 5.841497230529785,
      "epoch": 2.643765804318226,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042973362715146397,
      "loss": 5.1037,
      "mean_token_accuracy": 0.19736572206020356,
      "num_tokens": 58938148.0,
      "step": 33980
    },
    {
      "entropy": 5.753548002243042,
      "epoch": 2.6441548336899436,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004297135047316444,
      "loss": 4.9508,
      "mean_token_accuracy": 0.20714021921157838,
      "num_tokens": 58946406.0,
      "step": 33985
    },
    {
      "entropy": 5.771254062652588,
      "epoch": 2.644543863061661,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004296933799642936,
      "loss": 5.05,
      "mean_token_accuracy": 0.19927459061145783,
      "num_tokens": 58954914.0,
      "step": 33990
    },
    {
      "entropy": 5.7562675952911375,
      "epoch": 2.6449328924333786,
      "grad_norm": 1.34375,
      "learning_rate": 0.000429673252849717,
      "loss": 4.9793,
      "mean_token_accuracy": 0.20780300348997116,
      "num_tokens": 58963236.0,
      "step": 33995
    },
    {
      "entropy": 5.76298155784607,
      "epoch": 2.6453219218050963,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042965312338822005,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20675390362739562,
      "num_tokens": 58972141.0,
      "step": 34000
    },
    {
      "entropy": 5.756578969955444,
      "epoch": 2.645710951176814,
      "grad_norm": 1.109375,
      "learning_rate": 0.000429632991580108,
      "loss": 5.04,
      "mean_token_accuracy": 0.2009706199169159,
      "num_tokens": 58982348.0,
      "step": 34005
    },
    {
      "entropy": 5.804480981826782,
      "epoch": 2.6460999805485317,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004296128574256864,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20826094448566437,
      "num_tokens": 58990858.0,
      "step": 34010
    },
    {
      "entropy": 5.659407615661621,
      "epoch": 2.646489009920249,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042959272092526086,
      "loss": 4.8892,
      "mean_token_accuracy": 0.21642621457576752,
      "num_tokens": 58999841.0,
      "step": 34015
    },
    {
      "entropy": 5.687885332107544,
      "epoch": 2.6468780392919666,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042957258207913687,
      "loss": 4.921,
      "mean_token_accuracy": 0.20566900372505187,
      "num_tokens": 59008538.0,
      "step": 34020
    },
    {
      "entropy": 5.664876461029053,
      "epoch": 2.647267068663684,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042955244088761985,
      "loss": 4.9149,
      "mean_token_accuracy": 0.20497434139251708,
      "num_tokens": 59016656.0,
      "step": 34025
    },
    {
      "entropy": 5.750666522979737,
      "epoch": 2.6476560980354016,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004295322973510156,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21132152974605561,
      "num_tokens": 59025682.0,
      "step": 34030
    },
    {
      "entropy": 5.788916015625,
      "epoch": 2.6480451274071193,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004295121514696297,
      "loss": 5.0557,
      "mean_token_accuracy": 0.20646010786294938,
      "num_tokens": 59033448.0,
      "step": 34035
    },
    {
      "entropy": 5.7519886016845705,
      "epoch": 2.648434156778837,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042949200324376784,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21064203828573227,
      "num_tokens": 59042220.0,
      "step": 34040
    },
    {
      "entropy": 5.740723466873169,
      "epoch": 2.6488231861505542,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004294718526737357,
      "loss": 5.0413,
      "mean_token_accuracy": 0.2058459997177124,
      "num_tokens": 59050337.0,
      "step": 34045
    },
    {
      "entropy": 5.6967010498046875,
      "epoch": 2.649212215522272,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004294516997598391,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20317113399505615,
      "num_tokens": 59059186.0,
      "step": 34050
    },
    {
      "entropy": 5.783206224441528,
      "epoch": 2.6496012448939896,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004294315445023838,
      "loss": 5.0489,
      "mean_token_accuracy": 0.19469835758209228,
      "num_tokens": 59068714.0,
      "step": 34055
    },
    {
      "entropy": 5.878198337554932,
      "epoch": 2.649990274265707,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042941138690167556,
      "loss": 5.1319,
      "mean_token_accuracy": 0.19178012013435364,
      "num_tokens": 59077015.0,
      "step": 34060
    },
    {
      "entropy": 5.756258535385132,
      "epoch": 2.6503793036374246,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004293912269580204,
      "loss": 5.017,
      "mean_token_accuracy": 0.19916174560785294,
      "num_tokens": 59085766.0,
      "step": 34065
    },
    {
      "entropy": 5.829031038284302,
      "epoch": 2.6507683330091423,
      "grad_norm": 1.234375,
      "learning_rate": 0.000429371064671724,
      "loss": 5.0394,
      "mean_token_accuracy": 0.1949617862701416,
      "num_tokens": 59094390.0,
      "step": 34070
    },
    {
      "entropy": 5.8288524627685545,
      "epoch": 2.65115736238086,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042935090004309247,
      "loss": 5.0198,
      "mean_token_accuracy": 0.20847463011741638,
      "num_tokens": 59102310.0,
      "step": 34075
    },
    {
      "entropy": 5.788733673095703,
      "epoch": 2.6515463917525772,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042933073307243165,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20699029713869094,
      "num_tokens": 59111385.0,
      "step": 34080
    },
    {
      "entropy": 5.766987752914429,
      "epoch": 2.651935421124295,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042931056376004767,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21151384562253953,
      "num_tokens": 59120152.0,
      "step": 34085
    },
    {
      "entropy": 5.719979333877563,
      "epoch": 2.652324450496012,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004292903921062465,
      "loss": 4.8674,
      "mean_token_accuracy": 0.2192218005657196,
      "num_tokens": 59128782.0,
      "step": 34090
    },
    {
      "entropy": 5.7997535228729244,
      "epoch": 2.65271347986773,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042927021811133416,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20637050718069078,
      "num_tokens": 59136946.0,
      "step": 34095
    },
    {
      "entropy": 5.736763525009155,
      "epoch": 2.6531025092394476,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004292500417756167,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2125620126724243,
      "num_tokens": 59145845.0,
      "step": 34100
    },
    {
      "entropy": 5.827028560638428,
      "epoch": 2.6534915386111653,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042922986309940054,
      "loss": 5.0875,
      "mean_token_accuracy": 0.20092536211013795,
      "num_tokens": 59154607.0,
      "step": 34105
    },
    {
      "entropy": 5.724925708770752,
      "epoch": 2.653880567982883,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042920968208299165,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20646043568849565,
      "num_tokens": 59164160.0,
      "step": 34110
    },
    {
      "entropy": 5.818473672866821,
      "epoch": 2.6542695973546,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004291894987266963,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20885979384183884,
      "num_tokens": 59171976.0,
      "step": 34115
    },
    {
      "entropy": 5.737736511230469,
      "epoch": 2.654658626726318,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042916931303082064,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2087351992726326,
      "num_tokens": 59181168.0,
      "step": 34120
    },
    {
      "entropy": 5.813165712356567,
      "epoch": 2.655047656098035,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042914912499567113,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20063602775335312,
      "num_tokens": 59190047.0,
      "step": 34125
    },
    {
      "entropy": 5.791591930389404,
      "epoch": 2.655436685469753,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042912893462155395,
      "loss": 5.0186,
      "mean_token_accuracy": 0.20429133027791976,
      "num_tokens": 59199065.0,
      "step": 34130
    },
    {
      "entropy": 5.679439640045166,
      "epoch": 2.6558257148414706,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042910874190877545,
      "loss": 4.907,
      "mean_token_accuracy": 0.21661641150712968,
      "num_tokens": 59207827.0,
      "step": 34135
    },
    {
      "entropy": 5.786423873901367,
      "epoch": 2.6562147442131883,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004290885468576422,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1990243002772331,
      "num_tokens": 59216429.0,
      "step": 34140
    },
    {
      "entropy": 5.704759407043457,
      "epoch": 2.6566037735849055,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004290683494684604,
      "loss": 4.9703,
      "mean_token_accuracy": 0.1994563728570938,
      "num_tokens": 59224447.0,
      "step": 34145
    },
    {
      "entropy": 5.816413307189942,
      "epoch": 2.656992802956623,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004290481497415367,
      "loss": 5.0742,
      "mean_token_accuracy": 0.19399967342615126,
      "num_tokens": 59233124.0,
      "step": 34150
    },
    {
      "entropy": 5.727411365509033,
      "epoch": 2.657381832328341,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004290279476771775,
      "loss": 4.9346,
      "mean_token_accuracy": 0.20433439463376998,
      "num_tokens": 59241450.0,
      "step": 34155
    },
    {
      "entropy": 5.722670698165894,
      "epoch": 2.657770861700058,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004290077432756894,
      "loss": 5.1061,
      "mean_token_accuracy": 0.1943206235766411,
      "num_tokens": 59250797.0,
      "step": 34160
    },
    {
      "entropy": 5.667088603973388,
      "epoch": 2.658159891071776,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004289875365373788,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2028789237141609,
      "num_tokens": 59258854.0,
      "step": 34165
    },
    {
      "entropy": 5.776979637145996,
      "epoch": 2.6585489204434936,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042896732746255256,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20520206391811371,
      "num_tokens": 59267797.0,
      "step": 34170
    },
    {
      "entropy": 5.767652130126953,
      "epoch": 2.6589379498152113,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042894711605151714,
      "loss": 4.8953,
      "mean_token_accuracy": 0.21170506477355958,
      "num_tokens": 59276370.0,
      "step": 34175
    },
    {
      "entropy": 5.72540397644043,
      "epoch": 2.6593269791869285,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042892690230457924,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20508109033107758,
      "num_tokens": 59285641.0,
      "step": 34180
    },
    {
      "entropy": 5.72083854675293,
      "epoch": 2.659716008558646,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042890668622204566,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20930107533931733,
      "num_tokens": 59293816.0,
      "step": 34185
    },
    {
      "entropy": 5.7454986572265625,
      "epoch": 2.660105037930364,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000428886467804223,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21053119003772736,
      "num_tokens": 59301854.0,
      "step": 34190
    },
    {
      "entropy": 5.674050331115723,
      "epoch": 2.660494067302081,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042886624705141825,
      "loss": 4.9532,
      "mean_token_accuracy": 0.20742321908473968,
      "num_tokens": 59310798.0,
      "step": 34195
    },
    {
      "entropy": 5.758020496368408,
      "epoch": 2.660883096673799,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042884602396393796,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20321729630231858,
      "num_tokens": 59319085.0,
      "step": 34200
    },
    {
      "entropy": 5.742507648468018,
      "epoch": 2.6612721260455166,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004288257985420892,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20010122656822205,
      "num_tokens": 59327805.0,
      "step": 34205
    },
    {
      "entropy": 5.822107028961182,
      "epoch": 2.6616611554172342,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004288055707861788,
      "loss": 4.9627,
      "mean_token_accuracy": 0.20947305560112,
      "num_tokens": 59336520.0,
      "step": 34210
    },
    {
      "entropy": 5.732432460784912,
      "epoch": 2.6620501847889515,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042878534069651364,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20180297940969466,
      "num_tokens": 59345827.0,
      "step": 34215
    },
    {
      "entropy": 5.736692190170288,
      "epoch": 2.662439214160669,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004287651082734007,
      "loss": 5.0341,
      "mean_token_accuracy": 0.2076456904411316,
      "num_tokens": 59355024.0,
      "step": 34220
    },
    {
      "entropy": 5.711136531829834,
      "epoch": 2.6628282435323865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000428744873517147,
      "loss": 4.922,
      "mean_token_accuracy": 0.21254384368658066,
      "num_tokens": 59364461.0,
      "step": 34225
    },
    {
      "entropy": 5.768623495101929,
      "epoch": 2.663217272904104,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004287246364280596,
      "loss": 5.0841,
      "mean_token_accuracy": 0.19636242538690568,
      "num_tokens": 59372807.0,
      "step": 34230
    },
    {
      "entropy": 5.793723154067993,
      "epoch": 2.663606302275822,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004287043970064455,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20641451925039292,
      "num_tokens": 59381824.0,
      "step": 34235
    },
    {
      "entropy": 5.7328211784362795,
      "epoch": 2.6639953316475395,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004286841552526118,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21030170768499373,
      "num_tokens": 59391174.0,
      "step": 34240
    },
    {
      "entropy": 5.742779397964478,
      "epoch": 2.664384361019257,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042866391116686567,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2030511662364006,
      "num_tokens": 59399927.0,
      "step": 34245
    },
    {
      "entropy": 5.776315689086914,
      "epoch": 2.6647733903909745,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004286436647495142,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20864209085702895,
      "num_tokens": 59408016.0,
      "step": 34250
    },
    {
      "entropy": 5.73399863243103,
      "epoch": 2.665162419762692,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042862341600086477,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20175464898347856,
      "num_tokens": 59416702.0,
      "step": 34255
    },
    {
      "entropy": 5.756465196609497,
      "epoch": 2.6655514491344094,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042860316492122447,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2066696286201477,
      "num_tokens": 59424766.0,
      "step": 34260
    },
    {
      "entropy": 5.760356426239014,
      "epoch": 2.665940478506127,
      "grad_norm": 1.25,
      "learning_rate": 0.0004285829115109006,
      "loss": 5.0258,
      "mean_token_accuracy": 0.2074368804693222,
      "num_tokens": 59433691.0,
      "step": 34265
    },
    {
      "entropy": 5.793467473983765,
      "epoch": 2.666329507877845,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004285626557702005,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2141083836555481,
      "num_tokens": 59442495.0,
      "step": 34270
    },
    {
      "entropy": 5.742753267288208,
      "epoch": 2.6667185372495625,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004285423976994316,
      "loss": 4.9497,
      "mean_token_accuracy": 0.20887021869421005,
      "num_tokens": 59451038.0,
      "step": 34275
    },
    {
      "entropy": 5.665275573730469,
      "epoch": 2.66710756662128,
      "grad_norm": 1.25,
      "learning_rate": 0.00042852213729890117,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2081701472401619,
      "num_tokens": 59459119.0,
      "step": 34280
    },
    {
      "entropy": 5.732710170745849,
      "epoch": 2.6674965959929975,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004285018745689166,
      "loss": 4.9806,
      "mean_token_accuracy": 0.20428803265094758,
      "num_tokens": 59467725.0,
      "step": 34285
    },
    {
      "entropy": 5.776078367233277,
      "epoch": 2.667885625364715,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004284816095097855,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20458460003137588,
      "num_tokens": 59476231.0,
      "step": 34290
    },
    {
      "entropy": 5.703332901000977,
      "epoch": 2.6682746547364324,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004284613421218152,
      "loss": 4.9987,
      "mean_token_accuracy": 0.19833194017410277,
      "num_tokens": 59484898.0,
      "step": 34295
    },
    {
      "entropy": 5.691581344604492,
      "epoch": 2.66866368410815,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042844107240531324,
      "loss": 4.9326,
      "mean_token_accuracy": 0.20875302255153655,
      "num_tokens": 59493546.0,
      "step": 34300
    },
    {
      "entropy": 5.74297080039978,
      "epoch": 2.669052713479868,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004284208003605873,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20055067241191865,
      "num_tokens": 59502729.0,
      "step": 34305
    },
    {
      "entropy": 5.762607526779175,
      "epoch": 2.6694417428515855,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000428400525987945,
      "loss": 5.0386,
      "mean_token_accuracy": 0.20421160906553268,
      "num_tokens": 59512069.0,
      "step": 34310
    },
    {
      "entropy": 5.773636770248413,
      "epoch": 2.669830772223303,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042838024928769375,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20661787390708924,
      "num_tokens": 59520639.0,
      "step": 34315
    },
    {
      "entropy": 5.75873212814331,
      "epoch": 2.6702198015950205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042835997026014145,
      "loss": 4.927,
      "mean_token_accuracy": 0.2102395549416542,
      "num_tokens": 59528549.0,
      "step": 34320
    },
    {
      "entropy": 5.73044605255127,
      "epoch": 2.6706088309667377,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004283396889055957,
      "loss": 4.9528,
      "mean_token_accuracy": 0.21364280134439467,
      "num_tokens": 59537345.0,
      "step": 34325
    },
    {
      "entropy": 5.728127431869507,
      "epoch": 2.6709978603384554,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004283194052243641,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20486584901809693,
      "num_tokens": 59546866.0,
      "step": 34330
    },
    {
      "entropy": 5.7698217868804935,
      "epoch": 2.671386889710173,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042829911921675456,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20878684669733047,
      "num_tokens": 59555363.0,
      "step": 34335
    },
    {
      "entropy": 5.674425506591797,
      "epoch": 2.671775919081891,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000428278830883075,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21080251783132553,
      "num_tokens": 59564417.0,
      "step": 34340
    },
    {
      "entropy": 5.823581647872925,
      "epoch": 2.6721649484536085,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042825854022363307,
      "loss": 5.0017,
      "mean_token_accuracy": 0.20319636166095734,
      "num_tokens": 59572948.0,
      "step": 34345
    },
    {
      "entropy": 5.747403860092163,
      "epoch": 2.6725539778253258,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042823824723873676,
      "loss": 4.9409,
      "mean_token_accuracy": 0.20830875337123872,
      "num_tokens": 59581764.0,
      "step": 34350
    },
    {
      "entropy": 5.687945461273193,
      "epoch": 2.6729430071970435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042821795192869407,
      "loss": 4.9653,
      "mean_token_accuracy": 0.20999278575181962,
      "num_tokens": 59590281.0,
      "step": 34355
    },
    {
      "entropy": 5.802626514434815,
      "epoch": 2.6733320365687607,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004281976542938127,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1996137723326683,
      "num_tokens": 59598497.0,
      "step": 34360
    },
    {
      "entropy": 5.785196352005005,
      "epoch": 2.6737210659404784,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042817735433440076,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21007126420736313,
      "num_tokens": 59606901.0,
      "step": 34365
    },
    {
      "entropy": 5.831114768981934,
      "epoch": 2.674110095312196,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004281570520507663,
      "loss": 5.1172,
      "mean_token_accuracy": 0.20319176018238067,
      "num_tokens": 59616045.0,
      "step": 34370
    },
    {
      "entropy": 5.766650056838989,
      "epoch": 2.674499124683914,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042813674744321737,
      "loss": 5.0383,
      "mean_token_accuracy": 0.203377765417099,
      "num_tokens": 59624840.0,
      "step": 34375
    },
    {
      "entropy": 5.771958303451538,
      "epoch": 2.674888154055631,
      "grad_norm": 1.328125,
      "learning_rate": 0.000428116440512062,
      "loss": 5.019,
      "mean_token_accuracy": 0.19916785955429078,
      "num_tokens": 59632887.0,
      "step": 34380
    },
    {
      "entropy": 5.72641773223877,
      "epoch": 2.6752771834273488,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004280961312576084,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21060203313827514,
      "num_tokens": 59641671.0,
      "step": 34385
    },
    {
      "entropy": 5.686744308471679,
      "epoch": 2.6756662127990665,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004280758196801646,
      "loss": 4.8966,
      "mean_token_accuracy": 0.20980337709188462,
      "num_tokens": 59649406.0,
      "step": 34390
    },
    {
      "entropy": 5.817475366592407,
      "epoch": 2.6760552421707837,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042805550578003894,
      "loss": 5.0877,
      "mean_token_accuracy": 0.20021490305662154,
      "num_tokens": 59658114.0,
      "step": 34395
    },
    {
      "entropy": 5.752958297729492,
      "epoch": 2.6764442715425014,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004280351895575396,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20523583889007568,
      "num_tokens": 59667174.0,
      "step": 34400
    },
    {
      "entropy": 5.70434341430664,
      "epoch": 2.676833300914219,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004280148710129748,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20664681047201156,
      "num_tokens": 59675740.0,
      "step": 34405
    },
    {
      "entropy": 5.844768810272217,
      "epoch": 2.677222330285937,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042799455014665296,
      "loss": 5.1182,
      "mean_token_accuracy": 0.19936732947826385,
      "num_tokens": 59684550.0,
      "step": 34410
    },
    {
      "entropy": 5.84860634803772,
      "epoch": 2.677611359657654,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042797422695888226,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20311252474784852,
      "num_tokens": 59692810.0,
      "step": 34415
    },
    {
      "entropy": 5.742996072769165,
      "epoch": 2.6780003890293718,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004279539014499712,
      "loss": 4.9629,
      "mean_token_accuracy": 0.20220774561166763,
      "num_tokens": 59701742.0,
      "step": 34420
    },
    {
      "entropy": 5.731880855560303,
      "epoch": 2.678389418401089,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004279335736202281,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2025133416056633,
      "num_tokens": 59710440.0,
      "step": 34425
    },
    {
      "entropy": 5.848460054397583,
      "epoch": 2.6787784477728067,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004279132434699615,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20250306129455567,
      "num_tokens": 59720430.0,
      "step": 34430
    },
    {
      "entropy": 5.722204494476318,
      "epoch": 2.6791674771445244,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004278929109994798,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20641647279262543,
      "num_tokens": 59730249.0,
      "step": 34435
    },
    {
      "entropy": 5.716059684753418,
      "epoch": 2.679556506516242,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042787257620909144,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2071508377790451,
      "num_tokens": 59738815.0,
      "step": 34440
    },
    {
      "entropy": 5.672907209396362,
      "epoch": 2.67994553588796,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042785223909910515,
      "loss": 4.8895,
      "mean_token_accuracy": 0.20959192663431167,
      "num_tokens": 59747092.0,
      "step": 34445
    },
    {
      "entropy": 5.753254032135009,
      "epoch": 2.680334565259677,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004278318996698294,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20442702770233154,
      "num_tokens": 59756832.0,
      "step": 34450
    },
    {
      "entropy": 5.851410818099976,
      "epoch": 2.6807235946313948,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004278115579215728,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20116783678531647,
      "num_tokens": 59766052.0,
      "step": 34455
    },
    {
      "entropy": 5.74660096168518,
      "epoch": 2.681112624003112,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004277912138546441,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2100669801235199,
      "num_tokens": 59774369.0,
      "step": 34460
    },
    {
      "entropy": 5.699219846725464,
      "epoch": 2.6815016533748297,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004277708674693519,
      "loss": 4.9528,
      "mean_token_accuracy": 0.20918218940496444,
      "num_tokens": 59782509.0,
      "step": 34465
    },
    {
      "entropy": 5.65044584274292,
      "epoch": 2.6818906827465474,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042775051876600483,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2179492473602295,
      "num_tokens": 59790808.0,
      "step": 34470
    },
    {
      "entropy": 5.711146831512451,
      "epoch": 2.682279712118265,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004277301677449119,
      "loss": 4.8784,
      "mean_token_accuracy": 0.21040199398994447,
      "num_tokens": 59799260.0,
      "step": 34475
    },
    {
      "entropy": 5.764183521270752,
      "epoch": 2.6826687414899824,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004277098144063818,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20751554816961287,
      "num_tokens": 59807350.0,
      "step": 34480
    },
    {
      "entropy": 5.716615629196167,
      "epoch": 2.6830577708617,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042768945875072326,
      "loss": 5.038,
      "mean_token_accuracy": 0.19910196214914322,
      "num_tokens": 59816071.0,
      "step": 34485
    },
    {
      "entropy": 5.839780282974243,
      "epoch": 2.6834468002334178,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004276691007782451,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2117527589201927,
      "num_tokens": 59825332.0,
      "step": 34490
    },
    {
      "entropy": 5.767037677764892,
      "epoch": 2.683835829605135,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004276487404892565,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2016507253050804,
      "num_tokens": 59834082.0,
      "step": 34495
    },
    {
      "entropy": 5.768720817565918,
      "epoch": 2.6842248589768527,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042762837788406615,
      "loss": 5.0311,
      "mean_token_accuracy": 0.19867495000362395,
      "num_tokens": 59842633.0,
      "step": 34500
    },
    {
      "entropy": 5.808203220367432,
      "epoch": 2.6846138883485704,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004276080129629831,
      "loss": 4.9935,
      "mean_token_accuracy": 0.1987368106842041,
      "num_tokens": 59850517.0,
      "step": 34505
    },
    {
      "entropy": 5.812314176559449,
      "epoch": 2.685002917720288,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042758764572631644,
      "loss": 5.0868,
      "mean_token_accuracy": 0.20020653456449508,
      "num_tokens": 59859112.0,
      "step": 34510
    },
    {
      "entropy": 5.717612648010254,
      "epoch": 2.6853919470920053,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042756727617437504,
      "loss": 4.9139,
      "mean_token_accuracy": 0.21420759707689285,
      "num_tokens": 59868413.0,
      "step": 34515
    },
    {
      "entropy": 5.822158336639404,
      "epoch": 2.685780976463723,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042754690430746805,
      "loss": 5.0952,
      "mean_token_accuracy": 0.2021941512823105,
      "num_tokens": 59877300.0,
      "step": 34520
    },
    {
      "entropy": 5.734587526321411,
      "epoch": 2.6861700058354403,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004275265301259046,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2030343234539032,
      "num_tokens": 59885440.0,
      "step": 34525
    },
    {
      "entropy": 5.871235609054565,
      "epoch": 2.686559035207158,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042750615362999395,
      "loss": 5.0988,
      "mean_token_accuracy": 0.19646935760974885,
      "num_tokens": 59893954.0,
      "step": 34530
    },
    {
      "entropy": 5.758877420425415,
      "epoch": 2.6869480645788757,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042748577482004495,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20690529495477678,
      "num_tokens": 59902336.0,
      "step": 34535
    },
    {
      "entropy": 5.694613361358643,
      "epoch": 2.6873370939505934,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042746539369636715,
      "loss": 4.9021,
      "mean_token_accuracy": 0.21690521389245987,
      "num_tokens": 59911134.0,
      "step": 34540
    },
    {
      "entropy": 5.653019762039184,
      "epoch": 2.687726123322311,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042744501025926966,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21512297093868255,
      "num_tokens": 59920450.0,
      "step": 34545
    },
    {
      "entropy": 5.76284236907959,
      "epoch": 2.6881151526940283,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004274246245090618,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21495836824178696,
      "num_tokens": 59928919.0,
      "step": 34550
    },
    {
      "entropy": 5.75589337348938,
      "epoch": 2.688504182065746,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004274042364460529,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21316300183534623,
      "num_tokens": 59937973.0,
      "step": 34555
    },
    {
      "entropy": 5.745991516113281,
      "epoch": 2.6888932114374633,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042738384607055214,
      "loss": 5.012,
      "mean_token_accuracy": 0.20719306617975236,
      "num_tokens": 59946177.0,
      "step": 34560
    },
    {
      "entropy": 5.701706743240356,
      "epoch": 2.689282240809181,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042736345338286917,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20861773937940598,
      "num_tokens": 59955351.0,
      "step": 34565
    },
    {
      "entropy": 5.845680236816406,
      "epoch": 2.6896712701808987,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004273430583833133,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20133067816495895,
      "num_tokens": 59963839.0,
      "step": 34570
    },
    {
      "entropy": 5.802358770370484,
      "epoch": 2.6900602995526164,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000427322661072194,
      "loss": 5.0155,
      "mean_token_accuracy": 0.19529400318861007,
      "num_tokens": 59971779.0,
      "step": 34575
    },
    {
      "entropy": 5.775829839706421,
      "epoch": 2.6904493289243336,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004273022614498208,
      "loss": 4.9557,
      "mean_token_accuracy": 0.212911194562912,
      "num_tokens": 59980932.0,
      "step": 34580
    },
    {
      "entropy": 5.7570943355560305,
      "epoch": 2.6908383582960513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004272818595165031,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2128940239548683,
      "num_tokens": 59990281.0,
      "step": 34585
    },
    {
      "entropy": 5.687372922897339,
      "epoch": 2.691227387667769,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004272614552725506,
      "loss": 4.9092,
      "mean_token_accuracy": 0.215827514231205,
      "num_tokens": 59998642.0,
      "step": 34590
    },
    {
      "entropy": 5.753717279434204,
      "epoch": 2.6916164170394863,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042724104871827293,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2044452428817749,
      "num_tokens": 60006998.0,
      "step": 34595
    },
    {
      "entropy": 5.700570154190063,
      "epoch": 2.692005446411204,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004272206398539795,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21679009944200517,
      "num_tokens": 60015279.0,
      "step": 34600
    },
    {
      "entropy": 5.73796124458313,
      "epoch": 2.6923944757829217,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042720022867998026,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20709007531404494,
      "num_tokens": 60024648.0,
      "step": 34605
    },
    {
      "entropy": 5.802380514144898,
      "epoch": 2.6927835051546394,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004271798151965848,
      "loss": 5.0858,
      "mean_token_accuracy": 0.19976627677679062,
      "num_tokens": 60033821.0,
      "step": 34610
    },
    {
      "entropy": 5.754747486114502,
      "epoch": 2.6931725345263566,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004271593994041029,
      "loss": 4.989,
      "mean_token_accuracy": 0.20621783435344695,
      "num_tokens": 60041745.0,
      "step": 34615
    },
    {
      "entropy": 5.805905675888061,
      "epoch": 2.6935615638980743,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004271389813028443,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20435937047004699,
      "num_tokens": 60051659.0,
      "step": 34620
    },
    {
      "entropy": 5.851079416275025,
      "epoch": 2.693950593269792,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004271185608931187,
      "loss": 5.0881,
      "mean_token_accuracy": 0.19842259734869003,
      "num_tokens": 60060498.0,
      "step": 34625
    },
    {
      "entropy": 5.729318761825562,
      "epoch": 2.6943396226415093,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042709813817523617,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21109001338481903,
      "num_tokens": 60068678.0,
      "step": 34630
    },
    {
      "entropy": 5.723586702346802,
      "epoch": 2.694728652013227,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042707771314950645,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2091427445411682,
      "num_tokens": 60077889.0,
      "step": 34635
    },
    {
      "entropy": 5.75464506149292,
      "epoch": 2.6951176813849447,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042705728581623944,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2058095484972,
      "num_tokens": 60085989.0,
      "step": 34640
    },
    {
      "entropy": 5.7496050834655765,
      "epoch": 2.6955067107566624,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004270368561757452,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2073744058609009,
      "num_tokens": 60095012.0,
      "step": 34645
    },
    {
      "entropy": 5.778381252288819,
      "epoch": 2.6958957401283796,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004270164242283336,
      "loss": 5.0158,
      "mean_token_accuracy": 0.1960938423871994,
      "num_tokens": 60103894.0,
      "step": 34650
    },
    {
      "entropy": 5.683476161956787,
      "epoch": 2.6962847695000973,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004269959899743148,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21468450874090195,
      "num_tokens": 60112456.0,
      "step": 34655
    },
    {
      "entropy": 5.769298076629639,
      "epoch": 2.6966737988718146,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004269755534139987,
      "loss": 4.9984,
      "mean_token_accuracy": 0.20545771718025208,
      "num_tokens": 60120378.0,
      "step": 34660
    },
    {
      "entropy": 5.801659345626831,
      "epoch": 2.6970628282435323,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042695511454769557,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20882361978292466,
      "num_tokens": 60128091.0,
      "step": 34665
    },
    {
      "entropy": 5.77318320274353,
      "epoch": 2.69745185761525,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042693467337571534,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20769089609384536,
      "num_tokens": 60136298.0,
      "step": 34670
    },
    {
      "entropy": 5.804555511474609,
      "epoch": 2.6978408869869677,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004269142298983684,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21466342061758042,
      "num_tokens": 60145556.0,
      "step": 34675
    },
    {
      "entropy": 5.7864354133605955,
      "epoch": 2.698229916358685,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042689378411596467,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2089080974459648,
      "num_tokens": 60154301.0,
      "step": 34680
    },
    {
      "entropy": 5.651177310943604,
      "epoch": 2.6986189457304026,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004268733360288146,
      "loss": 4.917,
      "mean_token_accuracy": 0.22110872715711594,
      "num_tokens": 60162646.0,
      "step": 34685
    },
    {
      "entropy": 5.741998291015625,
      "epoch": 2.6990079751021203,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042685288563722835,
      "loss": 4.9868,
      "mean_token_accuracy": 0.204456128180027,
      "num_tokens": 60170749.0,
      "step": 34690
    },
    {
      "entropy": 5.731695508956909,
      "epoch": 2.6993970044738376,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004268324329415163,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21290068775415422,
      "num_tokens": 60178798.0,
      "step": 34695
    },
    {
      "entropy": 5.705108880996704,
      "epoch": 2.6997860338455553,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004268119779419887,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21131133288145065,
      "num_tokens": 60186974.0,
      "step": 34700
    },
    {
      "entropy": 5.689817905426025,
      "epoch": 2.700175063217273,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042679152063895603,
      "loss": 4.9313,
      "mean_token_accuracy": 0.20740904211997985,
      "num_tokens": 60195271.0,
      "step": 34705
    },
    {
      "entropy": 5.739785051345825,
      "epoch": 2.7005640925889907,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042677106103272867,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21037521064281464,
      "num_tokens": 60204476.0,
      "step": 34710
    },
    {
      "entropy": 5.77236967086792,
      "epoch": 2.700953121960708,
      "grad_norm": 1.21875,
      "learning_rate": 0.000426750599123617,
      "loss": 5.0255,
      "mean_token_accuracy": 0.20569379329681398,
      "num_tokens": 60213285.0,
      "step": 34715
    },
    {
      "entropy": 5.731132221221924,
      "epoch": 2.7013421513324256,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042673013491193145,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21160944551229477,
      "num_tokens": 60222321.0,
      "step": 34720
    },
    {
      "entropy": 5.7213136672973635,
      "epoch": 2.7017311807041433,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042670966839798263,
      "loss": 4.963,
      "mean_token_accuracy": 0.21830965280532838,
      "num_tokens": 60230759.0,
      "step": 34725
    },
    {
      "entropy": 5.77391996383667,
      "epoch": 2.7021202100758606,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004266891995820811,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20768416970968245,
      "num_tokens": 60238570.0,
      "step": 34730
    },
    {
      "entropy": 5.732962608337402,
      "epoch": 2.7025092394475783,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042666872846453734,
      "loss": 5.0232,
      "mean_token_accuracy": 0.2053869754076004,
      "num_tokens": 60247445.0,
      "step": 34735
    },
    {
      "entropy": 5.747666597366333,
      "epoch": 2.702898268819296,
      "grad_norm": 1.453125,
      "learning_rate": 0.00042664825504566206,
      "loss": 4.989,
      "mean_token_accuracy": 0.2015519395470619,
      "num_tokens": 60256358.0,
      "step": 34740
    },
    {
      "entropy": 5.710296249389648,
      "epoch": 2.7032872981910137,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000426627779325766,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20934805423021316,
      "num_tokens": 60264347.0,
      "step": 34745
    },
    {
      "entropy": 5.717221212387085,
      "epoch": 2.703676327562731,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004266073013051595,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20113002508878708,
      "num_tokens": 60273858.0,
      "step": 34750
    },
    {
      "entropy": 5.779476594924927,
      "epoch": 2.7040653569344486,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004265868209841537,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21410757005214692,
      "num_tokens": 60282089.0,
      "step": 34755
    },
    {
      "entropy": 5.740759944915771,
      "epoch": 2.704454386306166,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004265663383630591,
      "loss": 5.0297,
      "mean_token_accuracy": 0.2024807646870613,
      "num_tokens": 60291253.0,
      "step": 34760
    },
    {
      "entropy": 5.809834957122803,
      "epoch": 2.7048434156778836,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042654585344218657,
      "loss": 5.0123,
      "mean_token_accuracy": 0.20341800302267074,
      "num_tokens": 60300584.0,
      "step": 34765
    },
    {
      "entropy": 5.803416013717651,
      "epoch": 2.7052324450496013,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004265253662218469,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2064860939979553,
      "num_tokens": 60308822.0,
      "step": 34770
    },
    {
      "entropy": 5.771272039413452,
      "epoch": 2.705621474421319,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042650487670235093,
      "loss": 5.043,
      "mean_token_accuracy": 0.204637411236763,
      "num_tokens": 60317466.0,
      "step": 34775
    },
    {
      "entropy": 5.769265794754029,
      "epoch": 2.7060105037930366,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004264843848840097,
      "loss": 5.038,
      "mean_token_accuracy": 0.20233362466096877,
      "num_tokens": 60325307.0,
      "step": 34780
    },
    {
      "entropy": 5.835418558120727,
      "epoch": 2.706399533164754,
      "grad_norm": 1.21875,
      "learning_rate": 0.000426463890767134,
      "loss": 5.1046,
      "mean_token_accuracy": 0.19361608028411864,
      "num_tokens": 60334300.0,
      "step": 34785
    },
    {
      "entropy": 5.79053955078125,
      "epoch": 2.7067885625364716,
      "grad_norm": 1.375,
      "learning_rate": 0.0004264433943520349,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20314941555261612,
      "num_tokens": 60343481.0,
      "step": 34790
    },
    {
      "entropy": 5.74149374961853,
      "epoch": 2.707177591908189,
      "grad_norm": 1.125,
      "learning_rate": 0.00042642289563902326,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2033948227763176,
      "num_tokens": 60353134.0,
      "step": 34795
    },
    {
      "entropy": 5.7287139892578125,
      "epoch": 2.7075666212799065,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042640239462841023,
      "loss": 4.9398,
      "mean_token_accuracy": 0.2205089583992958,
      "num_tokens": 60362389.0,
      "step": 34800
    },
    {
      "entropy": 5.723274803161621,
      "epoch": 2.7079556506516242,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042638189132050675,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21340002566576005,
      "num_tokens": 60370963.0,
      "step": 34805
    },
    {
      "entropy": 5.7188890933990475,
      "epoch": 2.708344680023342,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042636138571562415,
      "loss": 5.0647,
      "mean_token_accuracy": 0.19698430746793746,
      "num_tokens": 60379534.0,
      "step": 34810
    },
    {
      "entropy": 5.774060487747192,
      "epoch": 2.708733709395059,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042634087781407337,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20257121026515962,
      "num_tokens": 60389234.0,
      "step": 34815
    },
    {
      "entropy": 5.784542846679687,
      "epoch": 2.709122738766777,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042632036761616564,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2064142107963562,
      "num_tokens": 60398571.0,
      "step": 34820
    },
    {
      "entropy": 5.776476716995239,
      "epoch": 2.7095117681384946,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042629985512221224,
      "loss": 5.0735,
      "mean_token_accuracy": 0.2009956419467926,
      "num_tokens": 60407356.0,
      "step": 34825
    },
    {
      "entropy": 5.804599952697754,
      "epoch": 2.709900797510212,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004262793403325243,
      "loss": 5.0914,
      "mean_token_accuracy": 0.2015853777527809,
      "num_tokens": 60415617.0,
      "step": 34830
    },
    {
      "entropy": 5.785373830795288,
      "epoch": 2.7102898268819295,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042625882324741316,
      "loss": 5.0442,
      "mean_token_accuracy": 0.2069658875465393,
      "num_tokens": 60424453.0,
      "step": 34835
    },
    {
      "entropy": 5.789905166625976,
      "epoch": 2.7106788562536472,
      "grad_norm": 1.25,
      "learning_rate": 0.0004262383038671902,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21438639611005783,
      "num_tokens": 60433358.0,
      "step": 34840
    },
    {
      "entropy": 5.726040363311768,
      "epoch": 2.711067885625365,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004262177821921666,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20344542860984802,
      "num_tokens": 60441793.0,
      "step": 34845
    },
    {
      "entropy": 5.760380601882934,
      "epoch": 2.711456914997082,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042619725822265394,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20971341282129288,
      "num_tokens": 60449981.0,
      "step": 34850
    },
    {
      "entropy": 5.747367191314697,
      "epoch": 2.7118459443688,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042617673195896345,
      "loss": 5.0525,
      "mean_token_accuracy": 0.19946722984313964,
      "num_tokens": 60459050.0,
      "step": 34855
    },
    {
      "entropy": 5.646168994903564,
      "epoch": 2.712234973740517,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042615620340140674,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2163206845521927,
      "num_tokens": 60467300.0,
      "step": 34860
    },
    {
      "entropy": 5.695121335983276,
      "epoch": 2.712624003112235,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042613567255029525,
      "loss": 4.9806,
      "mean_token_accuracy": 0.20772727727890014,
      "num_tokens": 60475969.0,
      "step": 34865
    },
    {
      "entropy": 5.724307632446289,
      "epoch": 2.7130130324839525,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042611513940594037,
      "loss": 4.9849,
      "mean_token_accuracy": 0.20683959126472473,
      "num_tokens": 60484886.0,
      "step": 34870
    },
    {
      "entropy": 5.7946713924407955,
      "epoch": 2.7134020618556702,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004260946039686539,
      "loss": 5.0399,
      "mean_token_accuracy": 0.20131461769342424,
      "num_tokens": 60494172.0,
      "step": 34875
    },
    {
      "entropy": 5.794255876541138,
      "epoch": 2.713791091227388,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042607406623874726,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20201028734445572,
      "num_tokens": 60501722.0,
      "step": 34880
    },
    {
      "entropy": 5.723868370056152,
      "epoch": 2.714180120599105,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004260535262165322,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21831660866737365,
      "num_tokens": 60510613.0,
      "step": 34885
    },
    {
      "entropy": 5.7594709396362305,
      "epoch": 2.714569149970823,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042603298390232015,
      "loss": 5.0626,
      "mean_token_accuracy": 0.20463392287492752,
      "num_tokens": 60519868.0,
      "step": 34890
    },
    {
      "entropy": 5.796018457412719,
      "epoch": 2.71495817934254,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042601243929642305,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20014730393886565,
      "num_tokens": 60528832.0,
      "step": 34895
    },
    {
      "entropy": 5.74968900680542,
      "epoch": 2.715347208714258,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042599189239915253,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20079136192798613,
      "num_tokens": 60537870.0,
      "step": 34900
    },
    {
      "entropy": 5.7468585014343265,
      "epoch": 2.7157362380859755,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004259713432108204,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2074611395597458,
      "num_tokens": 60546157.0,
      "step": 34905
    },
    {
      "entropy": 5.685452175140381,
      "epoch": 2.7161252674576932,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004259507917317384,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2034166172146797,
      "num_tokens": 60554411.0,
      "step": 34910
    },
    {
      "entropy": 5.82871994972229,
      "epoch": 2.7165142968294105,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042593023796221843,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20744359195232392,
      "num_tokens": 60562748.0,
      "step": 34915
    },
    {
      "entropy": 5.748939466476441,
      "epoch": 2.716903326201128,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004259096819025723,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21144218891859054,
      "num_tokens": 60572004.0,
      "step": 34920
    },
    {
      "entropy": 5.788563632965088,
      "epoch": 2.717292355572846,
      "grad_norm": 1.375,
      "learning_rate": 0.00042588912355311196,
      "loss": 5.0222,
      "mean_token_accuracy": 0.2050468549132347,
      "num_tokens": 60580596.0,
      "step": 34925
    },
    {
      "entropy": 5.753671598434448,
      "epoch": 2.717681384944563,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042586856291414925,
      "loss": 5.007,
      "mean_token_accuracy": 0.20499146431684495,
      "num_tokens": 60588734.0,
      "step": 34930
    },
    {
      "entropy": 5.767887544631958,
      "epoch": 2.718070414316281,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042584799998599633,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20750593543052673,
      "num_tokens": 60597302.0,
      "step": 34935
    },
    {
      "entropy": 5.79701976776123,
      "epoch": 2.7184594436879985,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004258274347689651,
      "loss": 5.0473,
      "mean_token_accuracy": 0.20208207964897157,
      "num_tokens": 60606928.0,
      "step": 34940
    },
    {
      "entropy": 5.781796026229858,
      "epoch": 2.718848473059716,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042580686726336766,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20767635405063628,
      "num_tokens": 60615970.0,
      "step": 34945
    },
    {
      "entropy": 5.748098516464234,
      "epoch": 2.7192375024314335,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042578629746951597,
      "loss": 5.0498,
      "mean_token_accuracy": 0.2032701164484024,
      "num_tokens": 60625453.0,
      "step": 34950
    },
    {
      "entropy": 5.7822660446167,
      "epoch": 2.719626531803151,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004257657253877222,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20794080197811127,
      "num_tokens": 60634622.0,
      "step": 34955
    },
    {
      "entropy": 5.7956702709198,
      "epoch": 2.720015561174869,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042574515101829856,
      "loss": 4.9904,
      "mean_token_accuracy": 0.2065092533826828,
      "num_tokens": 60643463.0,
      "step": 34960
    },
    {
      "entropy": 5.765991497039795,
      "epoch": 2.720404590546586,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042572457436155717,
      "loss": 4.9893,
      "mean_token_accuracy": 0.20614361017942429,
      "num_tokens": 60652192.0,
      "step": 34965
    },
    {
      "entropy": 5.85630578994751,
      "epoch": 2.720793619918304,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042570399541781023,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20122835785150528,
      "num_tokens": 60661071.0,
      "step": 34970
    },
    {
      "entropy": 5.865203142166138,
      "epoch": 2.7211826492900215,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042568341418737016,
      "loss": 5.1424,
      "mean_token_accuracy": 0.1928470626473427,
      "num_tokens": 60670165.0,
      "step": 34975
    },
    {
      "entropy": 5.757057952880859,
      "epoch": 2.721571678661739,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000425662830670549,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2084265097975731,
      "num_tokens": 60678260.0,
      "step": 34980
    },
    {
      "entropy": 5.755134534835816,
      "epoch": 2.7219607080334565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042564224486765925,
      "loss": 4.9446,
      "mean_token_accuracy": 0.20855761021375657,
      "num_tokens": 60686919.0,
      "step": 34985
    },
    {
      "entropy": 5.682869100570679,
      "epoch": 2.722349737405174,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004256216567790131,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21291155219078065,
      "num_tokens": 60695549.0,
      "step": 34990
    },
    {
      "entropy": 5.69507327079773,
      "epoch": 2.7227387667768914,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004256010664049232,
      "loss": 4.9492,
      "mean_token_accuracy": 0.202657350897789,
      "num_tokens": 60704258.0,
      "step": 34995
    },
    {
      "entropy": 5.780047655105591,
      "epoch": 2.723127796148609,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004255804737457017,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20294853001832963,
      "num_tokens": 60712860.0,
      "step": 35000
    },
    {
      "entropy": 5.775177335739135,
      "epoch": 2.723516825520327,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042555987880166127,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21291908174753188,
      "num_tokens": 60721861.0,
      "step": 35005
    },
    {
      "entropy": 5.776232957839966,
      "epoch": 2.7239058548920445,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004255392815731142,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21039292365312576,
      "num_tokens": 60730760.0,
      "step": 35010
    },
    {
      "entropy": 5.771334552764893,
      "epoch": 2.7242948842637618,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042551868206037324,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20172187089920043,
      "num_tokens": 60739192.0,
      "step": 35015
    },
    {
      "entropy": 5.7068703174591064,
      "epoch": 2.7246839136354795,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042549808026375084,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20665597319602966,
      "num_tokens": 60748009.0,
      "step": 35020
    },
    {
      "entropy": 5.741770267486572,
      "epoch": 2.725072943007197,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042547747618355966,
      "loss": 4.9935,
      "mean_token_accuracy": 0.20778180062770843,
      "num_tokens": 60756663.0,
      "step": 35025
    },
    {
      "entropy": 5.823272848129273,
      "epoch": 2.7254619723789144,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004254568698201122,
      "loss": 5.1453,
      "mean_token_accuracy": 0.19829111546278,
      "num_tokens": 60765055.0,
      "step": 35030
    },
    {
      "entropy": 5.785240983963012,
      "epoch": 2.725851001750632,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042543626117372133,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1987381175160408,
      "num_tokens": 60773745.0,
      "step": 35035
    },
    {
      "entropy": 5.790021562576294,
      "epoch": 2.72624003112235,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042541565024469945,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2069968417286873,
      "num_tokens": 60782244.0,
      "step": 35040
    },
    {
      "entropy": 5.70373797416687,
      "epoch": 2.7266290604940675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042539503703335963,
      "loss": 4.8636,
      "mean_token_accuracy": 0.21364473104476928,
      "num_tokens": 60791156.0,
      "step": 35045
    },
    {
      "entropy": 5.694369840621948,
      "epoch": 2.7270180898657848,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042537442154001455,
      "loss": 4.898,
      "mean_token_accuracy": 0.205828857421875,
      "num_tokens": 60799498.0,
      "step": 35050
    },
    {
      "entropy": 5.740624856948853,
      "epoch": 2.7274071192375025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042535380376497684,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1935053825378418,
      "num_tokens": 60807991.0,
      "step": 35055
    },
    {
      "entropy": 5.81803526878357,
      "epoch": 2.72779614860922,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004253331837085595,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20484774857759475,
      "num_tokens": 60815674.0,
      "step": 35060
    },
    {
      "entropy": 5.747046613693238,
      "epoch": 2.7281851779809374,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042531256137107545,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2105432242155075,
      "num_tokens": 60823490.0,
      "step": 35065
    },
    {
      "entropy": 5.727843427658081,
      "epoch": 2.728574207352655,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004252919367528375,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2064356043934822,
      "num_tokens": 60832437.0,
      "step": 35070
    },
    {
      "entropy": 5.746991300582886,
      "epoch": 2.728963236724373,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042527130985415857,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20213272869586946,
      "num_tokens": 60841450.0,
      "step": 35075
    },
    {
      "entropy": 5.7759300708770756,
      "epoch": 2.7293522660960905,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004252506806753517,
      "loss": 4.9353,
      "mean_token_accuracy": 0.21396380215883254,
      "num_tokens": 60849848.0,
      "step": 35080
    },
    {
      "entropy": 5.648829174041748,
      "epoch": 2.7297412954678077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042523004921673,
      "loss": 4.8881,
      "mean_token_accuracy": 0.20445971190929413,
      "num_tokens": 60858912.0,
      "step": 35085
    },
    {
      "entropy": 5.6834827899932865,
      "epoch": 2.7301303248395254,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004252094154786063,
      "loss": 4.9596,
      "mean_token_accuracy": 0.20626309961080552,
      "num_tokens": 60867204.0,
      "step": 35090
    },
    {
      "entropy": 5.722910356521607,
      "epoch": 2.7305193542112427,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004251887794612939,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20041292905807495,
      "num_tokens": 60875921.0,
      "step": 35095
    },
    {
      "entropy": 5.7330786228179935,
      "epoch": 2.7309083835829604,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004251681411651057,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20502367317676545,
      "num_tokens": 60884639.0,
      "step": 35100
    },
    {
      "entropy": 5.770031261444092,
      "epoch": 2.731297412954678,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042514750059035504,
      "loss": 5.1025,
      "mean_token_accuracy": 0.20553782284259797,
      "num_tokens": 60893517.0,
      "step": 35105
    },
    {
      "entropy": 5.666000938415527,
      "epoch": 2.731686442326396,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004251268577373551,
      "loss": 4.9781,
      "mean_token_accuracy": 0.2105904683470726,
      "num_tokens": 60902515.0,
      "step": 35110
    },
    {
      "entropy": 5.777387857437134,
      "epoch": 2.732075471698113,
      "grad_norm": 1.109375,
      "learning_rate": 0.000425106212606419,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20619254559278488,
      "num_tokens": 60911453.0,
      "step": 35115
    },
    {
      "entropy": 5.793831872940063,
      "epoch": 2.7324645010698307,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042508556519786004,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20915534049272538,
      "num_tokens": 60920549.0,
      "step": 35120
    },
    {
      "entropy": 5.813200187683106,
      "epoch": 2.7328535304415484,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004250649155119915,
      "loss": 5.0154,
      "mean_token_accuracy": 0.20657134503126146,
      "num_tokens": 60929271.0,
      "step": 35125
    },
    {
      "entropy": 5.765397262573242,
      "epoch": 2.7332425598132657,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042504426354912666,
      "loss": 4.9844,
      "mean_token_accuracy": 0.205043326318264,
      "num_tokens": 60938228.0,
      "step": 35130
    },
    {
      "entropy": 5.811329698562622,
      "epoch": 2.7336315891849834,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042502360930957904,
      "loss": 5.0349,
      "mean_token_accuracy": 0.2051067218184471,
      "num_tokens": 60945777.0,
      "step": 35135
    },
    {
      "entropy": 5.72904953956604,
      "epoch": 2.734020618556701,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042500295279366196,
      "loss": 5.0455,
      "mean_token_accuracy": 0.20099996626377106,
      "num_tokens": 60954500.0,
      "step": 35140
    },
    {
      "entropy": 5.757383823394775,
      "epoch": 2.734409647928419,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042498229400168874,
      "loss": 4.9953,
      "mean_token_accuracy": 0.20610959678888321,
      "num_tokens": 60962918.0,
      "step": 35145
    },
    {
      "entropy": 5.735591888427734,
      "epoch": 2.734798677300136,
      "grad_norm": 1.25,
      "learning_rate": 0.000424961632933973,
      "loss": 4.9601,
      "mean_token_accuracy": 0.20509911477565765,
      "num_tokens": 60971544.0,
      "step": 35150
    },
    {
      "entropy": 5.757019424438477,
      "epoch": 2.7351877066718537,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004249409695908281,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2071397617459297,
      "num_tokens": 60979637.0,
      "step": 35155
    },
    {
      "entropy": 5.6850159645080565,
      "epoch": 2.7355767360435714,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004249203039725677,
      "loss": 4.9966,
      "mean_token_accuracy": 0.19707981497049332,
      "num_tokens": 60988417.0,
      "step": 35160
    },
    {
      "entropy": 5.819476222991943,
      "epoch": 2.7359657654152887,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042489963607950527,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20674665421247482,
      "num_tokens": 60997282.0,
      "step": 35165
    },
    {
      "entropy": 5.799399471282959,
      "epoch": 2.7363547947870064,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042487896591195454,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20129559189081192,
      "num_tokens": 61006191.0,
      "step": 35170
    },
    {
      "entropy": 5.709873533248901,
      "epoch": 2.736743824158724,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042485829347022893,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21346058398485185,
      "num_tokens": 61014780.0,
      "step": 35175
    },
    {
      "entropy": 5.743605470657348,
      "epoch": 2.737132853530442,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042483761875464226,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20832220017910003,
      "num_tokens": 61024328.0,
      "step": 35180
    },
    {
      "entropy": 5.776262521743774,
      "epoch": 2.737521882902159,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042481694176550825,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21166949421167375,
      "num_tokens": 61032716.0,
      "step": 35185
    },
    {
      "entropy": 5.831011867523193,
      "epoch": 2.7379109122738767,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042479626250314066,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20450353771448135,
      "num_tokens": 61041495.0,
      "step": 35190
    },
    {
      "entropy": 5.702711534500122,
      "epoch": 2.738299941645594,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042477558096785316,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20488375574350357,
      "num_tokens": 61050275.0,
      "step": 35195
    },
    {
      "entropy": 5.714150333404541,
      "epoch": 2.7386889710173117,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042475489715995956,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22130033522844314,
      "num_tokens": 61058774.0,
      "step": 35200
    },
    {
      "entropy": 5.791710281372071,
      "epoch": 2.7390780003890294,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042473421107977385,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2126697540283203,
      "num_tokens": 61067475.0,
      "step": 35205
    },
    {
      "entropy": 5.727117824554443,
      "epoch": 2.739467029760747,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042471352272760974,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2126373380422592,
      "num_tokens": 61075943.0,
      "step": 35210
    },
    {
      "entropy": 5.763680934906006,
      "epoch": 2.7398560591324648,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004246928321037813,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2050071895122528,
      "num_tokens": 61084460.0,
      "step": 35215
    },
    {
      "entropy": 5.768786382675171,
      "epoch": 2.740245088504182,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004246721392086023,
      "loss": 5.0592,
      "mean_token_accuracy": 0.20791793912649154,
      "num_tokens": 61093423.0,
      "step": 35220
    },
    {
      "entropy": 5.770121145248413,
      "epoch": 2.7406341178758997,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042465144404238686,
      "loss": 5.0418,
      "mean_token_accuracy": 0.19923315048217774,
      "num_tokens": 61102953.0,
      "step": 35225
    },
    {
      "entropy": 5.767862129211426,
      "epoch": 2.741023147247617,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042463074660544897,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2177003026008606,
      "num_tokens": 61112114.0,
      "step": 35230
    },
    {
      "entropy": 5.816784286499024,
      "epoch": 2.7414121766193347,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004246100468981026,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21023241877555848,
      "num_tokens": 61120603.0,
      "step": 35235
    },
    {
      "entropy": 5.755855846405029,
      "epoch": 2.7418012059910524,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000424589344920662,
      "loss": 4.9759,
      "mean_token_accuracy": 0.20921741873025895,
      "num_tokens": 61129238.0,
      "step": 35240
    },
    {
      "entropy": 5.723650932312012,
      "epoch": 2.74219023536277,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004245686406734411,
      "loss": 4.8903,
      "mean_token_accuracy": 0.2181495890021324,
      "num_tokens": 61138082.0,
      "step": 35245
    },
    {
      "entropy": 5.80233325958252,
      "epoch": 2.7425792647344873,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004245479341567541,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20679081827402115,
      "num_tokens": 61146779.0,
      "step": 35250
    },
    {
      "entropy": 5.669378185272217,
      "epoch": 2.742968294106205,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004245272253709154,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2160259485244751,
      "num_tokens": 61155492.0,
      "step": 35255
    },
    {
      "entropy": 5.7062279224395756,
      "epoch": 2.7433573234779227,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004245065143162389,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21660805940628053,
      "num_tokens": 61164235.0,
      "step": 35260
    },
    {
      "entropy": 5.703236818313599,
      "epoch": 2.74374635284964,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042448580099303904,
      "loss": 4.9705,
      "mean_token_accuracy": 0.2086171180009842,
      "num_tokens": 61172203.0,
      "step": 35265
    },
    {
      "entropy": 5.740910148620605,
      "epoch": 2.7441353822213577,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042446508540163017,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20895354747772216,
      "num_tokens": 61180827.0,
      "step": 35270
    },
    {
      "entropy": 5.767629194259643,
      "epoch": 2.7445244115930754,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004244443675423264,
      "loss": 4.9927,
      "mean_token_accuracy": 0.20846739709377288,
      "num_tokens": 61189899.0,
      "step": 35275
    },
    {
      "entropy": 5.801315879821777,
      "epoch": 2.744913440964793,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042442364741544227,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1988482102751732,
      "num_tokens": 61197981.0,
      "step": 35280
    },
    {
      "entropy": 5.805603981018066,
      "epoch": 2.7453024703365103,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042440292502129214,
      "loss": 5.0073,
      "mean_token_accuracy": 0.2071913093328476,
      "num_tokens": 61206085.0,
      "step": 35285
    },
    {
      "entropy": 5.679562520980835,
      "epoch": 2.745691499708228,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004243822003601904,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20723973363637924,
      "num_tokens": 61215124.0,
      "step": 35290
    },
    {
      "entropy": 5.797019338607788,
      "epoch": 2.7460805290799453,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004243614734324515,
      "loss": 5.123,
      "mean_token_accuracy": 0.19796162098646164,
      "num_tokens": 61223926.0,
      "step": 35295
    },
    {
      "entropy": 5.747642755508423,
      "epoch": 2.746469558451663,
      "grad_norm": 1.25,
      "learning_rate": 0.0004243407442383901,
      "loss": 4.9786,
      "mean_token_accuracy": 0.20748327672481537,
      "num_tokens": 61232526.0,
      "step": 35300
    },
    {
      "entropy": 5.690400505065918,
      "epoch": 2.7468585878233807,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042432001277832045,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21566758453845977,
      "num_tokens": 61241591.0,
      "step": 35305
    },
    {
      "entropy": 5.7470029354095455,
      "epoch": 2.7472476171950984,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004242992790525573,
      "loss": 4.989,
      "mean_token_accuracy": 0.20870555490255355,
      "num_tokens": 61250851.0,
      "step": 35310
    },
    {
      "entropy": 5.760755348205566,
      "epoch": 2.747636646566816,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004242785430614154,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2131209135055542,
      "num_tokens": 61259600.0,
      "step": 35315
    },
    {
      "entropy": 5.742029237747192,
      "epoch": 2.7480256759385333,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042425780480520906,
      "loss": 5.0189,
      "mean_token_accuracy": 0.19889891594648362,
      "num_tokens": 61268458.0,
      "step": 35320
    },
    {
      "entropy": 5.718067169189453,
      "epoch": 2.748414705310251,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042423706428425305,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20448922514915466,
      "num_tokens": 61276974.0,
      "step": 35325
    },
    {
      "entropy": 5.841624546051025,
      "epoch": 2.7488037346819683,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004242163214988622,
      "loss": 5.0769,
      "mean_token_accuracy": 0.2041915625333786,
      "num_tokens": 61285165.0,
      "step": 35330
    },
    {
      "entropy": 5.767282962799072,
      "epoch": 2.749192764053686,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004241955764493512,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21373455673456193,
      "num_tokens": 61294061.0,
      "step": 35335
    },
    {
      "entropy": 5.74346399307251,
      "epoch": 2.7495817934254037,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042417482913603477,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21011377424001693,
      "num_tokens": 61302435.0,
      "step": 35340
    },
    {
      "entropy": 5.665187692642212,
      "epoch": 2.7499708227971214,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042415407955922774,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21077141910791397,
      "num_tokens": 61310462.0,
      "step": 35345
    },
    {
      "entropy": 5.800860404968262,
      "epoch": 2.7503598521688386,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042413332771924496,
      "loss": 5.1247,
      "mean_token_accuracy": 0.1877385050058365,
      "num_tokens": 61319490.0,
      "step": 35350
    },
    {
      "entropy": 5.883467483520508,
      "epoch": 2.7507488815405563,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004241125736164013,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20747675895690917,
      "num_tokens": 61328070.0,
      "step": 35355
    },
    {
      "entropy": 5.784058856964111,
      "epoch": 2.751137910912274,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004240918172510117,
      "loss": 4.8948,
      "mean_token_accuracy": 0.21431154608726502,
      "num_tokens": 61336319.0,
      "step": 35360
    },
    {
      "entropy": 5.718315982818604,
      "epoch": 2.7515269402839913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000424071058623391,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20586685538291932,
      "num_tokens": 61345564.0,
      "step": 35365
    },
    {
      "entropy": 5.75846848487854,
      "epoch": 2.751915969655709,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004240502977338543,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20485081523656845,
      "num_tokens": 61354823.0,
      "step": 35370
    },
    {
      "entropy": 5.82255482673645,
      "epoch": 2.7523049990274266,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004240295345827166,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20045577734708786,
      "num_tokens": 61363441.0,
      "step": 35375
    },
    {
      "entropy": 5.714589309692383,
      "epoch": 2.7526940283991443,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004240087691702929,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2031523510813713,
      "num_tokens": 61371795.0,
      "step": 35380
    },
    {
      "entropy": 5.736808204650879,
      "epoch": 2.7530830577708616,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004239880014968983,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2073907271027565,
      "num_tokens": 61379698.0,
      "step": 35385
    },
    {
      "entropy": 5.717116260528565,
      "epoch": 2.7534720871425793,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042396723156284786,
      "loss": 4.9442,
      "mean_token_accuracy": 0.2099240630865097,
      "num_tokens": 61388693.0,
      "step": 35390
    },
    {
      "entropy": 5.784937858581543,
      "epoch": 2.753861116514297,
      "grad_norm": 1.125,
      "learning_rate": 0.0004239464593684568,
      "loss": 5.029,
      "mean_token_accuracy": 0.20579161942005159,
      "num_tokens": 61397995.0,
      "step": 35395
    },
    {
      "entropy": 5.846347904205322,
      "epoch": 2.7542501458860142,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042392568491404043,
      "loss": 5.0712,
      "mean_token_accuracy": 0.2031252771615982,
      "num_tokens": 61405567.0,
      "step": 35400
    },
    {
      "entropy": 5.726482963562011,
      "epoch": 2.754639175257732,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042390490819991367,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21481054425239562,
      "num_tokens": 61413458.0,
      "step": 35405
    },
    {
      "entropy": 5.734873676300049,
      "epoch": 2.7550282046294496,
      "grad_norm": 1.296875,
      "learning_rate": 0.000423884129226392,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21470159143209458,
      "num_tokens": 61422026.0,
      "step": 35410
    },
    {
      "entropy": 5.832580089569092,
      "epoch": 2.7554172340011673,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004238633479937906,
      "loss": 5.0011,
      "mean_token_accuracy": 0.2102493017911911,
      "num_tokens": 61430843.0,
      "step": 35415
    },
    {
      "entropy": 5.716332960128784,
      "epoch": 2.7558062633728846,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004238425645024249,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21193906515836716,
      "num_tokens": 61438340.0,
      "step": 35420
    },
    {
      "entropy": 5.713162803649903,
      "epoch": 2.7561952927446023,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004238217787526102,
      "loss": 4.9507,
      "mean_token_accuracy": 0.20933496356010436,
      "num_tokens": 61447302.0,
      "step": 35425
    },
    {
      "entropy": 5.658839750289917,
      "epoch": 2.7565843221163195,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004238009907446618,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21639988124370574,
      "num_tokens": 61456688.0,
      "step": 35430
    },
    {
      "entropy": 5.846605825424194,
      "epoch": 2.7569733514880372,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042378020047889527,
      "loss": 5.0169,
      "mean_token_accuracy": 0.19691419452428818,
      "num_tokens": 61465718.0,
      "step": 35435
    },
    {
      "entropy": 5.86953182220459,
      "epoch": 2.757362380859755,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042375940795562594,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20243558287620544,
      "num_tokens": 61474113.0,
      "step": 35440
    },
    {
      "entropy": 5.70814323425293,
      "epoch": 2.7577514102314726,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042373861317516944,
      "loss": 4.947,
      "mean_token_accuracy": 0.20837783217430114,
      "num_tokens": 61482876.0,
      "step": 35445
    },
    {
      "entropy": 5.733007526397705,
      "epoch": 2.75814043960319,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004237178161378412,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2086677923798561,
      "num_tokens": 61491229.0,
      "step": 35450
    },
    {
      "entropy": 5.7043171405792235,
      "epoch": 2.7585294689749076,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004236970168439568,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20755314379930495,
      "num_tokens": 61499935.0,
      "step": 35455
    },
    {
      "entropy": 5.774708271026611,
      "epoch": 2.7589184983466253,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042367621529383186,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20812305957078933,
      "num_tokens": 61508249.0,
      "step": 35460
    },
    {
      "entropy": 5.760588264465332,
      "epoch": 2.7593075277183425,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042365541148778205,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2119321867823601,
      "num_tokens": 61516530.0,
      "step": 35465
    },
    {
      "entropy": 5.7844319343566895,
      "epoch": 2.7596965570900602,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042363460542612285,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2040475755929947,
      "num_tokens": 61524998.0,
      "step": 35470
    },
    {
      "entropy": 5.593046951293945,
      "epoch": 2.760085586461778,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004236137971091702,
      "loss": 4.8548,
      "mean_token_accuracy": 0.21654379814863206,
      "num_tokens": 61533661.0,
      "step": 35475
    },
    {
      "entropy": 5.742354154586792,
      "epoch": 2.7604746158334956,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042359298653723963,
      "loss": 4.9568,
      "mean_token_accuracy": 0.20942605882883072,
      "num_tokens": 61542387.0,
      "step": 35480
    },
    {
      "entropy": 5.755615377426148,
      "epoch": 2.760863645205213,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004235721737106471,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21658480167388916,
      "num_tokens": 61550626.0,
      "step": 35485
    },
    {
      "entropy": 5.7755042552948,
      "epoch": 2.7612526745769306,
      "grad_norm": 1.25,
      "learning_rate": 0.0004235513586297082,
      "loss": 5.0784,
      "mean_token_accuracy": 0.20433511435985566,
      "num_tokens": 61559170.0,
      "step": 35490
    },
    {
      "entropy": 5.780713081359863,
      "epoch": 2.7616417039486483,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042353054129473894,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2002916768193245,
      "num_tokens": 61567484.0,
      "step": 35495
    },
    {
      "entropy": 5.727398490905761,
      "epoch": 2.7620307333203655,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004235097217060552,
      "loss": 4.9183,
      "mean_token_accuracy": 0.21421697437763215,
      "num_tokens": 61575542.0,
      "step": 35500
    },
    {
      "entropy": 5.768709707260132,
      "epoch": 2.7624197626920832,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004234888998639727,
      "loss": 5.0371,
      "mean_token_accuracy": 0.20340640544891359,
      "num_tokens": 61584273.0,
      "step": 35505
    },
    {
      "entropy": 5.727620553970337,
      "epoch": 2.762808792063801,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004234680757688076,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2083084687590599,
      "num_tokens": 61593514.0,
      "step": 35510
    },
    {
      "entropy": 5.798971223831177,
      "epoch": 2.7631978214355186,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042344724942087564,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20328969955444337,
      "num_tokens": 61601146.0,
      "step": 35515
    },
    {
      "entropy": 5.789596319198608,
      "epoch": 2.763586850807236,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000423426420820493,
      "loss": 5.0586,
      "mean_token_accuracy": 0.2066815420985222,
      "num_tokens": 61609521.0,
      "step": 35520
    },
    {
      "entropy": 5.7682685375213625,
      "epoch": 2.7639758801789536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042340558996797564,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20049153566360473,
      "num_tokens": 61618382.0,
      "step": 35525
    },
    {
      "entropy": 5.740010786056518,
      "epoch": 2.764364909550671,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042338475686363964,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2110798954963684,
      "num_tokens": 61627098.0,
      "step": 35530
    },
    {
      "entropy": 5.739703178405762,
      "epoch": 2.7647539389223885,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004233639215078012,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20845955461263657,
      "num_tokens": 61635747.0,
      "step": 35535
    },
    {
      "entropy": 5.757989120483399,
      "epoch": 2.765142968294106,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042334308390077637,
      "loss": 5.0,
      "mean_token_accuracy": 0.2044748470187187,
      "num_tokens": 61645151.0,
      "step": 35540
    },
    {
      "entropy": 5.764014148712159,
      "epoch": 2.765531997665824,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004233222440428814,
      "loss": 5.0105,
      "mean_token_accuracy": 0.20802130550146103,
      "num_tokens": 61653216.0,
      "step": 35545
    },
    {
      "entropy": 5.796429586410523,
      "epoch": 2.765921027037541,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004233014019344324,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2098440483212471,
      "num_tokens": 61661583.0,
      "step": 35550
    },
    {
      "entropy": 5.714306735992432,
      "epoch": 2.766310056409259,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004232805575757457,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21066599637269973,
      "num_tokens": 61670465.0,
      "step": 35555
    },
    {
      "entropy": 5.7988440036773685,
      "epoch": 2.7666990857809766,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004232597109671375,
      "loss": 4.976,
      "mean_token_accuracy": 0.20788161605596542,
      "num_tokens": 61678670.0,
      "step": 35560
    },
    {
      "entropy": 5.841157865524292,
      "epoch": 2.767088115152694,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004232388621089242,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20262239426374434,
      "num_tokens": 61687860.0,
      "step": 35565
    },
    {
      "entropy": 5.77489161491394,
      "epoch": 2.7674771445244115,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042321801100142216,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20772851556539534,
      "num_tokens": 61696395.0,
      "step": 35570
    },
    {
      "entropy": 5.785155439376831,
      "epoch": 2.767866173896129,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042319715764494765,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20098426342010497,
      "num_tokens": 61705029.0,
      "step": 35575
    },
    {
      "entropy": 5.758699655532837,
      "epoch": 2.768255203267847,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042317630203981714,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2109498053789139,
      "num_tokens": 61713160.0,
      "step": 35580
    },
    {
      "entropy": 5.736287403106689,
      "epoch": 2.768644232639564,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042315544418634714,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21265812516212462,
      "num_tokens": 61722094.0,
      "step": 35585
    },
    {
      "entropy": 5.816694688796997,
      "epoch": 2.769033262011282,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042313458408485414,
      "loss": 5.0957,
      "mean_token_accuracy": 0.2039629653096199,
      "num_tokens": 61730771.0,
      "step": 35590
    },
    {
      "entropy": 5.749382591247558,
      "epoch": 2.7694222913829996,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00042311372173565456,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20609102249145508,
      "num_tokens": 61740059.0,
      "step": 35595
    },
    {
      "entropy": 5.791189289093017,
      "epoch": 2.769811320754717,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000423092857139065,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20565405786037444,
      "num_tokens": 61748466.0,
      "step": 35600
    },
    {
      "entropy": 5.7340058326721195,
      "epoch": 2.7702003501264345,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000423071990295402,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20764650404453278,
      "num_tokens": 61756862.0,
      "step": 35605
    },
    {
      "entropy": 5.72958459854126,
      "epoch": 2.770589379498152,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004230511212049822,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21179068833589554,
      "num_tokens": 61765867.0,
      "step": 35610
    },
    {
      "entropy": 5.754576301574707,
      "epoch": 2.77097840886987,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004230302498681224,
      "loss": 4.9786,
      "mean_token_accuracy": 0.20530669391155243,
      "num_tokens": 61774297.0,
      "step": 35615
    },
    {
      "entropy": 5.701286745071411,
      "epoch": 2.771367438241587,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004230093762851392,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21428552865982056,
      "num_tokens": 61782841.0,
      "step": 35620
    },
    {
      "entropy": 5.778935670852661,
      "epoch": 2.771756467613305,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004229885004563491,
      "loss": 5.1012,
      "mean_token_accuracy": 0.20414054542779922,
      "num_tokens": 61791719.0,
      "step": 35625
    },
    {
      "entropy": 5.802781867980957,
      "epoch": 2.772145496985022,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004229676223820692,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20308293551206588,
      "num_tokens": 61800267.0,
      "step": 35630
    },
    {
      "entropy": 5.749756336212158,
      "epoch": 2.77253452635674,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042294674206261616,
      "loss": 5.0151,
      "mean_token_accuracy": 0.2087683066725731,
      "num_tokens": 61808482.0,
      "step": 35635
    },
    {
      "entropy": 5.652102470397949,
      "epoch": 2.7729235557284575,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004229258594983067,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2175070509314537,
      "num_tokens": 61816950.0,
      "step": 35640
    },
    {
      "entropy": 5.730807209014893,
      "epoch": 2.773312585100175,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004229049746894578,
      "loss": 5.0158,
      "mean_token_accuracy": 0.20625619739294052,
      "num_tokens": 61826011.0,
      "step": 35645
    },
    {
      "entropy": 5.808103322982788,
      "epoch": 2.773701614471893,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004228840876363864,
      "loss": 5.0235,
      "mean_token_accuracy": 0.20174380540847778,
      "num_tokens": 61834705.0,
      "step": 35650
    },
    {
      "entropy": 5.795866537094116,
      "epoch": 2.77409064384361,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004228631983394093,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20565741658210754,
      "num_tokens": 61844166.0,
      "step": 35655
    },
    {
      "entropy": 5.799953746795654,
      "epoch": 2.774479673215328,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004228423067988434,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20677006393671035,
      "num_tokens": 61852141.0,
      "step": 35660
    },
    {
      "entropy": 5.74609489440918,
      "epoch": 2.774868702587045,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042282141301500597,
      "loss": 4.9997,
      "mean_token_accuracy": 0.2000180423259735,
      "num_tokens": 61861566.0,
      "step": 35665
    },
    {
      "entropy": 5.792853164672851,
      "epoch": 2.775257731958763,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004228005169882138,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20827166587114335,
      "num_tokens": 61869878.0,
      "step": 35670
    },
    {
      "entropy": 5.75975022315979,
      "epoch": 2.7756467613304805,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042277961871878413,
      "loss": 5.0177,
      "mean_token_accuracy": 0.19952283352613448,
      "num_tokens": 61879220.0,
      "step": 35675
    },
    {
      "entropy": 5.75281171798706,
      "epoch": 2.776035790702198,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004227587182070338,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21296288520097734,
      "num_tokens": 61887534.0,
      "step": 35680
    },
    {
      "entropy": 5.7774560928344725,
      "epoch": 2.7764248200739154,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004227378154532802,
      "loss": 5.1357,
      "mean_token_accuracy": 0.18909078389406203,
      "num_tokens": 61896420.0,
      "step": 35685
    },
    {
      "entropy": 5.77571439743042,
      "epoch": 2.776813849445633,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004227169104578404,
      "loss": 5.079,
      "mean_token_accuracy": 0.2043674647808075,
      "num_tokens": 61905024.0,
      "step": 35690
    },
    {
      "entropy": 5.706975317001342,
      "epoch": 2.777202878817351,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042269600322103164,
      "loss": 4.9674,
      "mean_token_accuracy": 0.20582322180271148,
      "num_tokens": 61913272.0,
      "step": 35695
    },
    {
      "entropy": 5.786889600753784,
      "epoch": 2.777591908189068,
      "grad_norm": 1.15625,
      "learning_rate": 0.000422675093743171,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20848263502120973,
      "num_tokens": 61922142.0,
      "step": 35700
    },
    {
      "entropy": 5.7523496627807615,
      "epoch": 2.777980937560786,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042265418202457603,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20928435623645783,
      "num_tokens": 61930140.0,
      "step": 35705
    },
    {
      "entropy": 5.76367564201355,
      "epoch": 2.7783699669325035,
      "grad_norm": 1.375,
      "learning_rate": 0.0004226332680655637,
      "loss": 5.0536,
      "mean_token_accuracy": 0.20453526824712753,
      "num_tokens": 61938993.0,
      "step": 35710
    },
    {
      "entropy": 5.751331090927124,
      "epoch": 2.778758996304221,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004226123518664516,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20340383648872376,
      "num_tokens": 61948063.0,
      "step": 35715
    },
    {
      "entropy": 5.756748247146606,
      "epoch": 2.7791480256759384,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004225914334275569,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20497924387454985,
      "num_tokens": 61956610.0,
      "step": 35720
    },
    {
      "entropy": 5.687800168991089,
      "epoch": 2.779537055047656,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004225705127491972,
      "loss": 4.8419,
      "mean_token_accuracy": 0.21141195744276048,
      "num_tokens": 61965359.0,
      "step": 35725
    },
    {
      "entropy": 5.717045545578003,
      "epoch": 2.7799260844193734,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004225495898316899,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20867631435394288,
      "num_tokens": 61973523.0,
      "step": 35730
    },
    {
      "entropy": 5.691704940795899,
      "epoch": 2.780315113791091,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042252866467535233,
      "loss": 4.9195,
      "mean_token_accuracy": 0.2164598286151886,
      "num_tokens": 61981988.0,
      "step": 35735
    },
    {
      "entropy": 5.7465736865997314,
      "epoch": 2.780704143162809,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042250773728050214,
      "loss": 4.9812,
      "mean_token_accuracy": 0.2099693164229393,
      "num_tokens": 61991400.0,
      "step": 35740
    },
    {
      "entropy": 5.749604415893555,
      "epoch": 2.7810931725345265,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004224868076474568,
      "loss": 5.059,
      "mean_token_accuracy": 0.20289977490901948,
      "num_tokens": 61999287.0,
      "step": 35745
    },
    {
      "entropy": 5.749526596069336,
      "epoch": 2.781482201906244,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042246587577653394,
      "loss": 5.006,
      "mean_token_accuracy": 0.20055113285779952,
      "num_tokens": 62008578.0,
      "step": 35750
    },
    {
      "entropy": 5.791533613204956,
      "epoch": 2.7818712312779614,
      "grad_norm": 1.25,
      "learning_rate": 0.0004224449416680511,
      "loss": 4.9272,
      "mean_token_accuracy": 0.21190676242113113,
      "num_tokens": 62017222.0,
      "step": 35755
    },
    {
      "entropy": 5.697207975387573,
      "epoch": 2.782260260649679,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042242400532232593,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21701498329639435,
      "num_tokens": 62026224.0,
      "step": 35760
    },
    {
      "entropy": 5.754910278320312,
      "epoch": 2.7826492900213964,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042240306673967614,
      "loss": 5.1516,
      "mean_token_accuracy": 0.1944774866104126,
      "num_tokens": 62036201.0,
      "step": 35765
    },
    {
      "entropy": 5.748770809173584,
      "epoch": 2.783038319393114,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042238212592041945,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21454015225172043,
      "num_tokens": 62044780.0,
      "step": 35770
    },
    {
      "entropy": 5.754393291473389,
      "epoch": 2.7834273487648318,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042236118286487364,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20715019404888152,
      "num_tokens": 62052720.0,
      "step": 35775
    },
    {
      "entropy": 5.7827911376953125,
      "epoch": 2.7838163781365495,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042234023757335633,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20114236772060395,
      "num_tokens": 62061797.0,
      "step": 35780
    },
    {
      "entropy": 5.73313274383545,
      "epoch": 2.7842054075082667,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042231929004618547,
      "loss": 4.9264,
      "mean_token_accuracy": 0.20873753875494003,
      "num_tokens": 62070543.0,
      "step": 35785
    },
    {
      "entropy": 5.823598575592041,
      "epoch": 2.7845944368799844,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004222983402836789,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20496804416179656,
      "num_tokens": 62079531.0,
      "step": 35790
    },
    {
      "entropy": 5.717537927627563,
      "epoch": 2.784983466251702,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042227738828615435,
      "loss": 4.9482,
      "mean_token_accuracy": 0.21194412261247636,
      "num_tokens": 62088429.0,
      "step": 35795
    },
    {
      "entropy": 5.714355373382569,
      "epoch": 2.7853724956234194,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042225643405392993,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21239645630121232,
      "num_tokens": 62096926.0,
      "step": 35800
    },
    {
      "entropy": 5.7544104099273685,
      "epoch": 2.785761524995137,
      "grad_norm": 1.25,
      "learning_rate": 0.00042223547758732356,
      "loss": 5.1126,
      "mean_token_accuracy": 0.19439491927623748,
      "num_tokens": 62106673.0,
      "step": 35805
    },
    {
      "entropy": 5.777977085113525,
      "epoch": 2.7861505543668548,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042221451888665307,
      "loss": 4.9697,
      "mean_token_accuracy": 0.20556211322546006,
      "num_tokens": 62115091.0,
      "step": 35810
    },
    {
      "entropy": 5.811440896987915,
      "epoch": 2.7865395837385725,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042219355795223657,
      "loss": 5.0189,
      "mean_token_accuracy": 0.20734891891479493,
      "num_tokens": 62124062.0,
      "step": 35815
    },
    {
      "entropy": 5.760553407669067,
      "epoch": 2.7869286131102897,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042217259478439216,
      "loss": 4.953,
      "mean_token_accuracy": 0.20766324251890184,
      "num_tokens": 62133141.0,
      "step": 35820
    },
    {
      "entropy": 5.7558910846710205,
      "epoch": 2.7873176424820074,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004221516293834379,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21007071435451508,
      "num_tokens": 62142056.0,
      "step": 35825
    },
    {
      "entropy": 5.8336587905883786,
      "epoch": 2.787706671853725,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042213066174969177,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20703255534172058,
      "num_tokens": 62150064.0,
      "step": 35830
    },
    {
      "entropy": 5.6931431770324705,
      "epoch": 2.7880957012254424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042210969188347217,
      "loss": 4.9308,
      "mean_token_accuracy": 0.20868754386901855,
      "num_tokens": 62159300.0,
      "step": 35835
    },
    {
      "entropy": 5.692614221572876,
      "epoch": 2.78848473059716,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004220887197850971,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2151864469051361,
      "num_tokens": 62167774.0,
      "step": 35840
    },
    {
      "entropy": 5.768594026565552,
      "epoch": 2.7888737599688778,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042206774545488476,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2115602523088455,
      "num_tokens": 62176301.0,
      "step": 35845
    },
    {
      "entropy": 5.750128126144409,
      "epoch": 2.7892627893405955,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004220467688931535,
      "loss": 4.9679,
      "mean_token_accuracy": 0.20708079636096954,
      "num_tokens": 62184804.0,
      "step": 35850
    },
    {
      "entropy": 5.755950307846069,
      "epoch": 2.7896518187123127,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004220257901002216,
      "loss": 4.9532,
      "mean_token_accuracy": 0.20789373964071273,
      "num_tokens": 62193864.0,
      "step": 35855
    },
    {
      "entropy": 5.754052734375,
      "epoch": 2.7900408480840304,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004220048090764073,
      "loss": 4.9689,
      "mean_token_accuracy": 0.20796546190977097,
      "num_tokens": 62202906.0,
      "step": 35860
    },
    {
      "entropy": 5.66084771156311,
      "epoch": 2.7904298774557477,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042198382582202905,
      "loss": 4.8407,
      "mean_token_accuracy": 0.21798142343759536,
      "num_tokens": 62211292.0,
      "step": 35865
    },
    {
      "entropy": 5.765640449523926,
      "epoch": 2.7908189068274654,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004219628403374052,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20127335339784622,
      "num_tokens": 62220830.0,
      "step": 35870
    },
    {
      "entropy": 5.7653178691864015,
      "epoch": 2.791207936199183,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004219418526228541,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2082146555185318,
      "num_tokens": 62228659.0,
      "step": 35875
    },
    {
      "entropy": 5.816193962097168,
      "epoch": 2.7915969655709008,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004219208626786943,
      "loss": 5.0884,
      "mean_token_accuracy": 0.19511766731739044,
      "num_tokens": 62238098.0,
      "step": 35880
    },
    {
      "entropy": 5.82524037361145,
      "epoch": 2.791985994942618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004218998705052443,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19710233807563782,
      "num_tokens": 62247523.0,
      "step": 35885
    },
    {
      "entropy": 5.780804061889649,
      "epoch": 2.7923750243143357,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042187887610282255,
      "loss": 4.9862,
      "mean_token_accuracy": 0.20078454166650772,
      "num_tokens": 62255601.0,
      "step": 35890
    },
    {
      "entropy": 5.783425235748291,
      "epoch": 2.7927640536860534,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004218578794717476,
      "loss": 4.9968,
      "mean_token_accuracy": 0.20645197331905366,
      "num_tokens": 62263650.0,
      "step": 35895
    },
    {
      "entropy": 5.728555393218994,
      "epoch": 2.7931530830577707,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042183688061233813,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20604331642389298,
      "num_tokens": 62272474.0,
      "step": 35900
    },
    {
      "entropy": 5.771803522109986,
      "epoch": 2.7935421124294884,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042181587952491267,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21951699256896973,
      "num_tokens": 62281364.0,
      "step": 35905
    },
    {
      "entropy": 5.807551097869873,
      "epoch": 2.793931141801206,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042179487620978997,
      "loss": 4.956,
      "mean_token_accuracy": 0.2065717965364456,
      "num_tokens": 62289025.0,
      "step": 35910
    },
    {
      "entropy": 5.718358039855957,
      "epoch": 2.7943201711729238,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004217738706672886,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2099643364548683,
      "num_tokens": 62298063.0,
      "step": 35915
    },
    {
      "entropy": 5.738404655456543,
      "epoch": 2.794709200544641,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042175286289772746,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21020981222391127,
      "num_tokens": 62306806.0,
      "step": 35920
    },
    {
      "entropy": 5.697347974777221,
      "epoch": 2.7950982299163587,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004217318529014251,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2052266016602516,
      "num_tokens": 62315119.0,
      "step": 35925
    },
    {
      "entropy": 5.757594156265259,
      "epoch": 2.7954872592880764,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004217108406787005,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2043006420135498,
      "num_tokens": 62323615.0,
      "step": 35930
    },
    {
      "entropy": 5.786683750152588,
      "epoch": 2.7958762886597937,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004216898262298724,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21374136060476304,
      "num_tokens": 62332313.0,
      "step": 35935
    },
    {
      "entropy": 5.651415586471558,
      "epoch": 2.7962653180315113,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004216688095552596,
      "loss": 4.8456,
      "mean_token_accuracy": 0.21206575781106948,
      "num_tokens": 62341060.0,
      "step": 35940
    },
    {
      "entropy": 5.7888020992279055,
      "epoch": 2.796654347403229,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004216477906551811,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20431183725595475,
      "num_tokens": 62349616.0,
      "step": 35945
    },
    {
      "entropy": 5.769074392318726,
      "epoch": 2.7970433767749467,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042162676952995585,
      "loss": 5.006,
      "mean_token_accuracy": 0.20727517902851106,
      "num_tokens": 62358083.0,
      "step": 35950
    },
    {
      "entropy": 5.796360111236572,
      "epoch": 2.797432406146664,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042160574617990267,
      "loss": 5.0068,
      "mean_token_accuracy": 0.20007237941026687,
      "num_tokens": 62366494.0,
      "step": 35955
    },
    {
      "entropy": 5.75521936416626,
      "epoch": 2.7978214355183817,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042158472060534066,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2030385911464691,
      "num_tokens": 62375590.0,
      "step": 35960
    },
    {
      "entropy": 5.787105751037598,
      "epoch": 2.798210464890099,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042156369280658886,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20545031428337096,
      "num_tokens": 62384186.0,
      "step": 35965
    },
    {
      "entropy": 5.796723651885986,
      "epoch": 2.7985994942618166,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004215426627839662,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20822616964578627,
      "num_tokens": 62392789.0,
      "step": 35970
    },
    {
      "entropy": 5.763696908950806,
      "epoch": 2.7989885236335343,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042152163053779196,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20779064446687698,
      "num_tokens": 62401808.0,
      "step": 35975
    },
    {
      "entropy": 5.79776177406311,
      "epoch": 2.799377553005252,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004215005960683852,
      "loss": 5.0279,
      "mean_token_accuracy": 0.20308367758989335,
      "num_tokens": 62410700.0,
      "step": 35980
    },
    {
      "entropy": 5.75486192703247,
      "epoch": 2.7997665823769697,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004214795593760651,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21660100817680358,
      "num_tokens": 62419337.0,
      "step": 35985
    },
    {
      "entropy": 5.758059787750244,
      "epoch": 2.800155611748687,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004214585204611508,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20014241635799407,
      "num_tokens": 62427189.0,
      "step": 35990
    },
    {
      "entropy": 5.77078685760498,
      "epoch": 2.8005446411204047,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042143747932396153,
      "loss": 5.082,
      "mean_token_accuracy": 0.2028965324163437,
      "num_tokens": 62436531.0,
      "step": 35995
    },
    {
      "entropy": 5.8213143825531,
      "epoch": 2.800933670492122,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004214164359648166,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20544271022081376,
      "num_tokens": 62445250.0,
      "step": 36000
    },
    {
      "epoch": 2.800933670492122,
      "eval_entropy": 5.544179245964395,
      "eval_loss": 5.070761203765869,
      "eval_mean_token_accuracy": 0.21278336730367065,
      "eval_num_tokens": 62445250.0,
      "eval_runtime": 22.0884,
      "eval_samples_per_second": 1881.442,
      "eval_steps_per_second": 235.192,
      "step": 36000
    },
    {
      "entropy": 5.7605678081512455,
      "epoch": 2.8013226998638396,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042139539038403535,
      "loss": 4.949,
      "mean_token_accuracy": 0.20130173712968827,
      "num_tokens": 62454221.0,
      "step": 36005
    },
    {
      "entropy": 5.684249830245972,
      "epoch": 2.8017117292355573,
      "grad_norm": 1.25,
      "learning_rate": 0.000421374342581937,
      "loss": 4.8763,
      "mean_token_accuracy": 0.21784111112356186,
      "num_tokens": 62463237.0,
      "step": 36010
    },
    {
      "entropy": 5.726301956176758,
      "epoch": 2.802100758607275,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042135329255884105,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21145346909761428,
      "num_tokens": 62472227.0,
      "step": 36015
    },
    {
      "entropy": 5.719448804855347,
      "epoch": 2.8024897879789923,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004213322403150668,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2089378297328949,
      "num_tokens": 62480225.0,
      "step": 36020
    },
    {
      "entropy": 5.758980560302734,
      "epoch": 2.80287881735071,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004213111858509337,
      "loss": 5.019,
      "mean_token_accuracy": 0.20261331051588058,
      "num_tokens": 62487858.0,
      "step": 36025
    },
    {
      "entropy": 5.803186845779419,
      "epoch": 2.8032678467224277,
      "grad_norm": 1.25,
      "learning_rate": 0.0004212901291667612,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19521222859621049,
      "num_tokens": 62496004.0,
      "step": 36030
    },
    {
      "entropy": 5.750979328155518,
      "epoch": 2.803656876094145,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042126907026286887,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20110173225402833,
      "num_tokens": 62504011.0,
      "step": 36035
    },
    {
      "entropy": 5.806132030487061,
      "epoch": 2.8040459054658626,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042124800913957624,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20178937613964082,
      "num_tokens": 62512143.0,
      "step": 36040
    },
    {
      "entropy": 5.8470104217529295,
      "epoch": 2.8044349348375803,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004212269457972028,
      "loss": 4.998,
      "mean_token_accuracy": 0.20512158870697023,
      "num_tokens": 62520374.0,
      "step": 36045
    },
    {
      "entropy": 5.86097207069397,
      "epoch": 2.804823964209298,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004212058802360682,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19382400661706925,
      "num_tokens": 62529345.0,
      "step": 36050
    },
    {
      "entropy": 5.698241329193115,
      "epoch": 2.8052129935810153,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004211848124564922,
      "loss": 4.9618,
      "mean_token_accuracy": 0.20744539052248,
      "num_tokens": 62538622.0,
      "step": 36055
    },
    {
      "entropy": 5.761926746368408,
      "epoch": 2.805602022952733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004211637424587942,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21148415207862853,
      "num_tokens": 62547071.0,
      "step": 36060
    },
    {
      "entropy": 5.821884059906006,
      "epoch": 2.8059910523244502,
      "grad_norm": 1.375,
      "learning_rate": 0.0004211426702432941,
      "loss": 5.0309,
      "mean_token_accuracy": 0.2013270989060402,
      "num_tokens": 62556321.0,
      "step": 36065
    },
    {
      "entropy": 5.807388114929199,
      "epoch": 2.806380081696168,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004211215958103116,
      "loss": 5.1395,
      "mean_token_accuracy": 0.2007797747850418,
      "num_tokens": 62565549.0,
      "step": 36070
    },
    {
      "entropy": 5.686150074005127,
      "epoch": 2.8067691110678856,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042110051916016645,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21293116807937623,
      "num_tokens": 62573753.0,
      "step": 36075
    },
    {
      "entropy": 5.7812947750091555,
      "epoch": 2.8071581404396033,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004210794402931785,
      "loss": 5.0326,
      "mean_token_accuracy": 0.19944216310977936,
      "num_tokens": 62583919.0,
      "step": 36080
    },
    {
      "entropy": 5.754004716873169,
      "epoch": 2.807547169811321,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004210583592096675,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2053047835826874,
      "num_tokens": 62592136.0,
      "step": 36085
    },
    {
      "entropy": 5.749445676803589,
      "epoch": 2.8079361991830383,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004210372759099534,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2115714132785797,
      "num_tokens": 62600742.0,
      "step": 36090
    },
    {
      "entropy": 5.771989345550537,
      "epoch": 2.808325228554756,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000421016190394356,
      "loss": 4.937,
      "mean_token_accuracy": 0.20871618539094924,
      "num_tokens": 62608239.0,
      "step": 36095
    },
    {
      "entropy": 5.782424068450927,
      "epoch": 2.8087142579264732,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042099510266319545,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20787828117609025,
      "num_tokens": 62617445.0,
      "step": 36100
    },
    {
      "entropy": 5.738164234161377,
      "epoch": 2.809103287298191,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004209740127167915,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20695659667253494,
      "num_tokens": 62625578.0,
      "step": 36105
    },
    {
      "entropy": 5.725192880630493,
      "epoch": 2.8094923166699086,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004209529205554643,
      "loss": 4.9539,
      "mean_token_accuracy": 0.20795880854129792,
      "num_tokens": 62635028.0,
      "step": 36110
    },
    {
      "entropy": 5.786328458786011,
      "epoch": 2.8098813460416263,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004209318261795339,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20402650386095048,
      "num_tokens": 62643739.0,
      "step": 36115
    },
    {
      "entropy": 5.80153226852417,
      "epoch": 2.8102703754133436,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004209107295893202,
      "loss": 5.1012,
      "mean_token_accuracy": 0.2043169841170311,
      "num_tokens": 62652769.0,
      "step": 36120
    },
    {
      "entropy": 5.712349510192871,
      "epoch": 2.8106594047850613,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042088963078514344,
      "loss": 4.989,
      "mean_token_accuracy": 0.20022280365228654,
      "num_tokens": 62661412.0,
      "step": 36125
    },
    {
      "entropy": 5.728562927246093,
      "epoch": 2.811048434156779,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004208685297673238,
      "loss": 5.0156,
      "mean_token_accuracy": 0.20508743673563004,
      "num_tokens": 62670048.0,
      "step": 36130
    },
    {
      "entropy": 5.812819576263427,
      "epoch": 2.811437463528496,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004208474265361813,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21149763762950896,
      "num_tokens": 62678901.0,
      "step": 36135
    },
    {
      "entropy": 5.744204998016357,
      "epoch": 2.811826492900214,
      "grad_norm": 1.25,
      "learning_rate": 0.00042082632109203627,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2157478079199791,
      "num_tokens": 62687124.0,
      "step": 36140
    },
    {
      "entropy": 5.777745246887207,
      "epoch": 2.8122155222719316,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000420805213435209,
      "loss": 5.04,
      "mean_token_accuracy": 0.2022998198866844,
      "num_tokens": 62696029.0,
      "step": 36145
    },
    {
      "entropy": 5.742279767990112,
      "epoch": 2.8126045516436493,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004207841035660196,
      "loss": 4.9032,
      "mean_token_accuracy": 0.20972758680582046,
      "num_tokens": 62704181.0,
      "step": 36150
    },
    {
      "entropy": 5.74112229347229,
      "epoch": 2.8129935810153666,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004207629914847885,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20455042868852616,
      "num_tokens": 62713440.0,
      "step": 36155
    },
    {
      "entropy": 5.747869873046875,
      "epoch": 2.8133826103870843,
      "grad_norm": 1.234375,
      "learning_rate": 0.000420741877191836,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20880212634801865,
      "num_tokens": 62723024.0,
      "step": 36160
    },
    {
      "entropy": 5.711106967926026,
      "epoch": 2.8137716397588015,
      "grad_norm": 1.25,
      "learning_rate": 0.0004207207606874825,
      "loss": 4.9926,
      "mean_token_accuracy": 0.20495866239070892,
      "num_tokens": 62731441.0,
      "step": 36165
    },
    {
      "entropy": 5.773100137710571,
      "epoch": 2.814160669130519,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004206996419720484,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20250794142484665,
      "num_tokens": 62740073.0,
      "step": 36170
    },
    {
      "entropy": 5.706322383880615,
      "epoch": 2.814549698502237,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004206785210458541,
      "loss": 4.8744,
      "mean_token_accuracy": 0.21897899359464645,
      "num_tokens": 62749192.0,
      "step": 36175
    },
    {
      "entropy": 5.795069837570191,
      "epoch": 2.8149387278739546,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004206573979092202,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21418891847133636,
      "num_tokens": 62757408.0,
      "step": 36180
    },
    {
      "entropy": 5.731108236312866,
      "epoch": 2.8153277572456723,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042063627256246706,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20317787528038025,
      "num_tokens": 62767475.0,
      "step": 36185
    },
    {
      "entropy": 5.660362958908081,
      "epoch": 2.8157167866173896,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004206151450059153,
      "loss": 4.8786,
      "mean_token_accuracy": 0.21493865549564362,
      "num_tokens": 62775802.0,
      "step": 36190
    },
    {
      "entropy": 5.658012104034424,
      "epoch": 2.8161058159891073,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042059401523988546,
      "loss": 4.942,
      "mean_token_accuracy": 0.20602052956819533,
      "num_tokens": 62784718.0,
      "step": 36195
    },
    {
      "entropy": 5.726744842529297,
      "epoch": 2.8164948453608245,
      "grad_norm": 1.25,
      "learning_rate": 0.0004205728832646982,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20026170015335082,
      "num_tokens": 62793692.0,
      "step": 36200
    },
    {
      "entropy": 5.780411863327027,
      "epoch": 2.816883874732542,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042055174908067414,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20334072709083556,
      "num_tokens": 62802583.0,
      "step": 36205
    },
    {
      "entropy": 5.794972658157349,
      "epoch": 2.81727290410426,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042053061268813397,
      "loss": 4.9795,
      "mean_token_accuracy": 0.2095896452665329,
      "num_tokens": 62812374.0,
      "step": 36210
    },
    {
      "entropy": 5.764403867721557,
      "epoch": 2.8176619334759776,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042050947408739836,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20237126499414443,
      "num_tokens": 62821796.0,
      "step": 36215
    },
    {
      "entropy": 5.731749439239502,
      "epoch": 2.818050962847695,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004204883332787881,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20623738169670106,
      "num_tokens": 62831349.0,
      "step": 36220
    },
    {
      "entropy": 5.7830055236816404,
      "epoch": 2.8184399922194125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042046719026262396,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21239063143730164,
      "num_tokens": 62838842.0,
      "step": 36225
    },
    {
      "entropy": 5.737091207504273,
      "epoch": 2.8188290215911302,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042044604503922674,
      "loss": 4.9509,
      "mean_token_accuracy": 0.20879091024398805,
      "num_tokens": 62847067.0,
      "step": 36230
    },
    {
      "entropy": 5.670303726196289,
      "epoch": 2.8192180509628475,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004204248976089173,
      "loss": 4.8501,
      "mean_token_accuracy": 0.2154502674937248,
      "num_tokens": 62855374.0,
      "step": 36235
    },
    {
      "entropy": 5.722654724121094,
      "epoch": 2.819607080334565,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004204037479720165,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21029646694660187,
      "num_tokens": 62863141.0,
      "step": 36240
    },
    {
      "entropy": 5.756218290328979,
      "epoch": 2.819996109706283,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042038259612884514,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20480761975049971,
      "num_tokens": 62872080.0,
      "step": 36245
    },
    {
      "entropy": 5.766063451766968,
      "epoch": 2.8203851390780006,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042036144207972445,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20569838732481002,
      "num_tokens": 62880492.0,
      "step": 36250
    },
    {
      "entropy": 5.754472160339356,
      "epoch": 2.820774168449718,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004203402858249752,
      "loss": 4.9494,
      "mean_token_accuracy": 0.20678077191114425,
      "num_tokens": 62888961.0,
      "step": 36255
    },
    {
      "entropy": 5.755246877670288,
      "epoch": 2.8211631978214355,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004203191273649184,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21131541281938554,
      "num_tokens": 62897224.0,
      "step": 36260
    },
    {
      "entropy": 5.73275146484375,
      "epoch": 2.8215522271931532,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042029796669987513,
      "loss": 4.9874,
      "mean_token_accuracy": 0.20293329656124115,
      "num_tokens": 62905873.0,
      "step": 36265
    },
    {
      "entropy": 5.746466445922851,
      "epoch": 2.8219412565648705,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004202768038301665,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2081998884677887,
      "num_tokens": 62914438.0,
      "step": 36270
    },
    {
      "entropy": 5.809631824493408,
      "epoch": 2.822330285936588,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042025563875611367,
      "loss": 5.0338,
      "mean_token_accuracy": 0.19971662014722824,
      "num_tokens": 62923310.0,
      "step": 36275
    },
    {
      "entropy": 5.773028945922851,
      "epoch": 2.822719315308306,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004202344714780376,
      "loss": 5.0707,
      "mean_token_accuracy": 0.2039431154727936,
      "num_tokens": 62932031.0,
      "step": 36280
    },
    {
      "entropy": 5.773457002639771,
      "epoch": 2.8231083446800236,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042021330199625966,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20650348365306853,
      "num_tokens": 62939695.0,
      "step": 36285
    },
    {
      "entropy": 5.743734216690063,
      "epoch": 2.823497374051741,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042019213031110105,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20961772799491882,
      "num_tokens": 62948251.0,
      "step": 36290
    },
    {
      "entropy": 5.78606767654419,
      "epoch": 2.8238864034234585,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004201709564228829,
      "loss": 4.9973,
      "mean_token_accuracy": 0.19952250868082047,
      "num_tokens": 62956881.0,
      "step": 36295
    },
    {
      "entropy": 5.905514287948608,
      "epoch": 2.824275432795176,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042014978033192655,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19304392039775847,
      "num_tokens": 62966241.0,
      "step": 36300
    },
    {
      "entropy": 5.730713891983032,
      "epoch": 2.8246644621668935,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042012860203855337,
      "loss": 4.8892,
      "mean_token_accuracy": 0.20843609869480134,
      "num_tokens": 62974826.0,
      "step": 36305
    },
    {
      "entropy": 5.80407862663269,
      "epoch": 2.825053491538611,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042010742154308453,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20281084030866622,
      "num_tokens": 62983800.0,
      "step": 36310
    },
    {
      "entropy": 5.7725952625274655,
      "epoch": 2.825442520910329,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042008623884584167,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20330837070941926,
      "num_tokens": 62993219.0,
      "step": 36315
    },
    {
      "entropy": 5.795035171508789,
      "epoch": 2.825831550282046,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042006505394714597,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20878974497318267,
      "num_tokens": 63002455.0,
      "step": 36320
    },
    {
      "entropy": 5.828357696533203,
      "epoch": 2.826220579653764,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042004386684731896,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19695955365896226,
      "num_tokens": 63012200.0,
      "step": 36325
    },
    {
      "entropy": 5.7894261360168455,
      "epoch": 2.8266096090254815,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042002267754668216,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2112940013408661,
      "num_tokens": 63021870.0,
      "step": 36330
    },
    {
      "entropy": 5.77707462310791,
      "epoch": 2.826998638397199,
      "grad_norm": 1.28125,
      "learning_rate": 0.000420001486045557,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2041461646556854,
      "num_tokens": 63030941.0,
      "step": 36335
    },
    {
      "entropy": 5.812815761566162,
      "epoch": 2.8273876677689165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004199802923442652,
      "loss": 5.045,
      "mean_token_accuracy": 0.20099622309207915,
      "num_tokens": 63039264.0,
      "step": 36340
    },
    {
      "entropy": 5.79822473526001,
      "epoch": 2.827776697140634,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004199590964431282,
      "loss": 5.0404,
      "mean_token_accuracy": 0.1996802046895027,
      "num_tokens": 63048723.0,
      "step": 36345
    },
    {
      "entropy": 5.773813056945801,
      "epoch": 2.828165726512352,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004199378983424676,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2079397216439247,
      "num_tokens": 63058147.0,
      "step": 36350
    },
    {
      "entropy": 5.774773693084716,
      "epoch": 2.828554755884069,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041991669804260505,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20991783887147902,
      "num_tokens": 63066614.0,
      "step": 36355
    },
    {
      "entropy": 5.762890291213989,
      "epoch": 2.828943785255787,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004198954955438623,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20469450801610947,
      "num_tokens": 63075174.0,
      "step": 36360
    },
    {
      "entropy": 5.783041524887085,
      "epoch": 2.8293328146275045,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000419874290846561,
      "loss": 4.9819,
      "mean_token_accuracy": 0.20862242728471755,
      "num_tokens": 63083375.0,
      "step": 36365
    },
    {
      "entropy": 5.76266303062439,
      "epoch": 2.8297218439992218,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041985308395102303,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21503864526748656,
      "num_tokens": 63091669.0,
      "step": 36370
    },
    {
      "entropy": 5.7157494068145756,
      "epoch": 2.8301108733709395,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041983187485756997,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21002158373594285,
      "num_tokens": 63100597.0,
      "step": 36375
    },
    {
      "entropy": 5.829881572723389,
      "epoch": 2.830499902742657,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041981066356652373,
      "loss": 5.1026,
      "mean_token_accuracy": 0.2011214718222618,
      "num_tokens": 63109445.0,
      "step": 36380
    },
    {
      "entropy": 5.806922769546508,
      "epoch": 2.830888932114375,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041978945007820626,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20112878531217576,
      "num_tokens": 63118485.0,
      "step": 36385
    },
    {
      "entropy": 5.8002214431762695,
      "epoch": 2.831277961486092,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041976823439293925,
      "loss": 4.9976,
      "mean_token_accuracy": 0.2089894950389862,
      "num_tokens": 63127366.0,
      "step": 36390
    },
    {
      "entropy": 5.730260133743286,
      "epoch": 2.83166699085781,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004197470165110447,
      "loss": 5.0299,
      "mean_token_accuracy": 0.2066662147641182,
      "num_tokens": 63136192.0,
      "step": 36395
    },
    {
      "entropy": 5.861303329467773,
      "epoch": 2.832056020229527,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004197257964328446,
      "loss": 5.1008,
      "mean_token_accuracy": 0.2008321389555931,
      "num_tokens": 63144250.0,
      "step": 36400
    },
    {
      "entropy": 5.810700130462647,
      "epoch": 2.8324450496012448,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004197045741586609,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20706386715173722,
      "num_tokens": 63152685.0,
      "step": 36405
    },
    {
      "entropy": 5.775984621047973,
      "epoch": 2.8328340789729625,
      "grad_norm": 1.25,
      "learning_rate": 0.0004196833496888156,
      "loss": 4.927,
      "mean_token_accuracy": 0.21140723675489426,
      "num_tokens": 63161853.0,
      "step": 36410
    },
    {
      "entropy": 5.681947183609009,
      "epoch": 2.83322310834468,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004196621230236308,
      "loss": 4.9457,
      "mean_token_accuracy": 0.2096162036061287,
      "num_tokens": 63170626.0,
      "step": 36415
    },
    {
      "entropy": 5.731553840637207,
      "epoch": 2.833612137716398,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041964089416342845,
      "loss": 5.0023,
      "mean_token_accuracy": 0.20221182107925414,
      "num_tokens": 63180183.0,
      "step": 36420
    },
    {
      "entropy": 5.8273491859436035,
      "epoch": 2.834001167088115,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004196196631085308,
      "loss": 5.0271,
      "mean_token_accuracy": 0.19924837499856948,
      "num_tokens": 63188993.0,
      "step": 36425
    },
    {
      "entropy": 5.790616989135742,
      "epoch": 2.834390196459833,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041959842985925987,
      "loss": 5.043,
      "mean_token_accuracy": 0.2000345006585121,
      "num_tokens": 63197592.0,
      "step": 36430
    },
    {
      "entropy": 5.746353816986084,
      "epoch": 2.83477922583155,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000419577194415938,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20369576215744017,
      "num_tokens": 63206216.0,
      "step": 36435
    },
    {
      "entropy": 5.722529602050781,
      "epoch": 2.8351682552032678,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041955595677888735,
      "loss": 4.9585,
      "mean_token_accuracy": 0.20760246813297273,
      "num_tokens": 63214633.0,
      "step": 36440
    },
    {
      "entropy": 5.738190412521362,
      "epoch": 2.8355572845749855,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004195347169484301,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2075570747256279,
      "num_tokens": 63223666.0,
      "step": 36445
    },
    {
      "entropy": 5.723755836486816,
      "epoch": 2.835946313946703,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004195134749248886,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21733589470386505,
      "num_tokens": 63232192.0,
      "step": 36450
    },
    {
      "entropy": 5.698632383346558,
      "epoch": 2.8363353433184204,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004194922307085851,
      "loss": 4.9692,
      "mean_token_accuracy": 0.20126767903566362,
      "num_tokens": 63241235.0,
      "step": 36455
    },
    {
      "entropy": 5.815583610534668,
      "epoch": 2.836724372690138,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004194709842998419,
      "loss": 5.0722,
      "mean_token_accuracy": 0.20011547803878785,
      "num_tokens": 63249742.0,
      "step": 36460
    },
    {
      "entropy": 5.741321086883545,
      "epoch": 2.837113402061856,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041944973569898157,
      "loss": 4.9669,
      "mean_token_accuracy": 0.20503271520137786,
      "num_tokens": 63258004.0,
      "step": 36465
    },
    {
      "entropy": 5.8229039192199705,
      "epoch": 2.837502431433573,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004194284849063265,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20148445963859557,
      "num_tokens": 63266478.0,
      "step": 36470
    },
    {
      "entropy": 5.7547869205474855,
      "epoch": 2.8378914608052908,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004194072319221989,
      "loss": 4.9818,
      "mean_token_accuracy": 0.20639038681983948,
      "num_tokens": 63274868.0,
      "step": 36475
    },
    {
      "entropy": 5.761697673797608,
      "epoch": 2.8382804901770085,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004193859767469214,
      "loss": 5.0932,
      "mean_token_accuracy": 0.1944795474410057,
      "num_tokens": 63284207.0,
      "step": 36480
    },
    {
      "entropy": 5.786730670928955,
      "epoch": 2.838669519548726,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041936471938081667,
      "loss": 5.0512,
      "mean_token_accuracy": 0.19900553524494172,
      "num_tokens": 63292477.0,
      "step": 36485
    },
    {
      "entropy": 5.752643156051636,
      "epoch": 2.8390585489204434,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041934345982420695,
      "loss": 5.0198,
      "mean_token_accuracy": 0.20232656747102737,
      "num_tokens": 63301457.0,
      "step": 36490
    },
    {
      "entropy": 5.847551965713501,
      "epoch": 2.839447578292161,
      "grad_norm": 1.296875,
      "learning_rate": 0.000419322198077415,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2018572449684143,
      "num_tokens": 63309967.0,
      "step": 36495
    },
    {
      "entropy": 5.774679327011109,
      "epoch": 2.8398366076638784,
      "grad_norm": 1.25,
      "learning_rate": 0.00041930093414076344,
      "loss": 4.8917,
      "mean_token_accuracy": 0.21216070502996445,
      "num_tokens": 63319329.0,
      "step": 36500
    },
    {
      "entropy": 5.699241685867309,
      "epoch": 2.840225637035596,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041927966801457486,
      "loss": 5.002,
      "mean_token_accuracy": 0.2096702516078949,
      "num_tokens": 63328600.0,
      "step": 36505
    },
    {
      "entropy": 5.776415157318115,
      "epoch": 2.8406146664073137,
      "grad_norm": 1.1875,
      "learning_rate": 0.000419258399699172,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21248516738414763,
      "num_tokens": 63337297.0,
      "step": 36510
    },
    {
      "entropy": 5.7575233459472654,
      "epoch": 2.8410036957790314,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004192371291948775,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2077943876385689,
      "num_tokens": 63346634.0,
      "step": 36515
    },
    {
      "entropy": 5.847502899169922,
      "epoch": 2.841392725150749,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041921585650201413,
      "loss": 5.0649,
      "mean_token_accuracy": 0.2000417962670326,
      "num_tokens": 63354870.0,
      "step": 36520
    },
    {
      "entropy": 5.736571264266968,
      "epoch": 2.8417817545224664,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004191945816209047,
      "loss": 5.045,
      "mean_token_accuracy": 0.19826662838459014,
      "num_tokens": 63363571.0,
      "step": 36525
    },
    {
      "entropy": 5.780000114440918,
      "epoch": 2.842170783894184,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041917330455187195,
      "loss": 4.9613,
      "mean_token_accuracy": 0.20939005315303802,
      "num_tokens": 63371716.0,
      "step": 36530
    },
    {
      "entropy": 5.730573272705078,
      "epoch": 2.8425598132659013,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041915202529523894,
      "loss": 5.0614,
      "mean_token_accuracy": 0.20208433419466018,
      "num_tokens": 63381399.0,
      "step": 36535
    },
    {
      "entropy": 5.794855785369873,
      "epoch": 2.842948842637619,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004191307438513283,
      "loss": 5.046,
      "mean_token_accuracy": 0.20782616138458251,
      "num_tokens": 63390548.0,
      "step": 36540
    },
    {
      "entropy": 5.751839113235474,
      "epoch": 2.8433378720093367,
      "grad_norm": 1.265625,
      "learning_rate": 0.000419109460220463,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20630524009466172,
      "num_tokens": 63400219.0,
      "step": 36545
    },
    {
      "entropy": 5.731627655029297,
      "epoch": 2.8437269013810544,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041908817440296605,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21356104165315629,
      "num_tokens": 63409131.0,
      "step": 36550
    },
    {
      "entropy": 5.800327777862549,
      "epoch": 2.8441159307527717,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041906688639916035,
      "loss": 5.1376,
      "mean_token_accuracy": 0.19569737911224366,
      "num_tokens": 63418830.0,
      "step": 36555
    },
    {
      "entropy": 5.777229928970337,
      "epoch": 2.8445049601244894,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004190455962093689,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20374457389116288,
      "num_tokens": 63427467.0,
      "step": 36560
    },
    {
      "entropy": 5.785231828689575,
      "epoch": 2.844893989496207,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00041902430383391494,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20537900179624557,
      "num_tokens": 63436957.0,
      "step": 36565
    },
    {
      "entropy": 5.764430379867553,
      "epoch": 2.8452830188679243,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004190030092731214,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20617408752441407,
      "num_tokens": 63446010.0,
      "step": 36570
    },
    {
      "entropy": 5.727068376541138,
      "epoch": 2.845672048239642,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004189817125273113,
      "loss": 4.931,
      "mean_token_accuracy": 0.20961880385875703,
      "num_tokens": 63454666.0,
      "step": 36575
    },
    {
      "entropy": 5.738446474075317,
      "epoch": 2.8460610776113597,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041896041359680797,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21437811553478242,
      "num_tokens": 63463076.0,
      "step": 36580
    },
    {
      "entropy": 5.771926641464233,
      "epoch": 2.8464501069830774,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041893911248193437,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2012687772512436,
      "num_tokens": 63471547.0,
      "step": 36585
    },
    {
      "entropy": 5.782917404174805,
      "epoch": 2.8468391363547947,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000418917809183014,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20373743623495102,
      "num_tokens": 63479710.0,
      "step": 36590
    },
    {
      "entropy": 5.718512105941772,
      "epoch": 2.8472281657265124,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041889650370036986,
      "loss": 4.9513,
      "mean_token_accuracy": 0.20786044448614122,
      "num_tokens": 63488639.0,
      "step": 36595
    },
    {
      "entropy": 5.793718528747559,
      "epoch": 2.84761719509823,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004188751960343253,
      "loss": 5.0725,
      "mean_token_accuracy": 0.19747330546379088,
      "num_tokens": 63496956.0,
      "step": 36600
    },
    {
      "entropy": 5.792957019805908,
      "epoch": 2.8480062244699473,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004188538861852037,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21026264429092406,
      "num_tokens": 63505538.0,
      "step": 36605
    },
    {
      "entropy": 5.691541290283203,
      "epoch": 2.848395253841665,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004188325741533283,
      "loss": 4.8793,
      "mean_token_accuracy": 0.21030011624097825,
      "num_tokens": 63513664.0,
      "step": 36610
    },
    {
      "entropy": 5.691984701156616,
      "epoch": 2.8487842832133827,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004188112599390225,
      "loss": 4.8792,
      "mean_token_accuracy": 0.21970844119787217,
      "num_tokens": 63521910.0,
      "step": 36615
    },
    {
      "entropy": 5.7292139530181885,
      "epoch": 2.8491733125851004,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004187899435426098,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21145769506692885,
      "num_tokens": 63530272.0,
      "step": 36620
    },
    {
      "entropy": 5.767516946792602,
      "epoch": 2.8495623419568177,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041876862496441347,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20535948723554612,
      "num_tokens": 63538652.0,
      "step": 36625
    },
    {
      "entropy": 5.792075967788696,
      "epoch": 2.8499513713285354,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041874730420475716,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20936208218336105,
      "num_tokens": 63547016.0,
      "step": 36630
    },
    {
      "entropy": 5.729784202575684,
      "epoch": 2.8503404007002526,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041872598126396434,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20970568507909776,
      "num_tokens": 63555561.0,
      "step": 36635
    },
    {
      "entropy": 5.723135185241699,
      "epoch": 2.8507294300719703,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041870465614235843,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20080848783254623,
      "num_tokens": 63563510.0,
      "step": 36640
    },
    {
      "entropy": 5.748687839508056,
      "epoch": 2.851118459443688,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041868332884026317,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2121207073330879,
      "num_tokens": 63572495.0,
      "step": 36645
    },
    {
      "entropy": 5.815672063827515,
      "epoch": 2.8515074888154057,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004186619993580021,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20385617911815643,
      "num_tokens": 63581525.0,
      "step": 36650
    },
    {
      "entropy": 5.7502867698669435,
      "epoch": 2.851896518187123,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041864066769589875,
      "loss": 4.9525,
      "mean_token_accuracy": 0.20684794038534166,
      "num_tokens": 63590136.0,
      "step": 36655
    },
    {
      "entropy": 5.798794841766357,
      "epoch": 2.8522855475588407,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041861933385427697,
      "loss": 5.0799,
      "mean_token_accuracy": 0.20363539308309556,
      "num_tokens": 63599353.0,
      "step": 36660
    },
    {
      "entropy": 5.733447456359864,
      "epoch": 2.8526745769305584,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004185979978334604,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21526931077241898,
      "num_tokens": 63607586.0,
      "step": 36665
    },
    {
      "entropy": 5.718154525756836,
      "epoch": 2.8530636063022756,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004185766596337727,
      "loss": 4.9367,
      "mean_token_accuracy": 0.20965086668729782,
      "num_tokens": 63616444.0,
      "step": 36670
    },
    {
      "entropy": 5.72216305732727,
      "epoch": 2.8534526356739933,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041855531925553773,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20277995020151138,
      "num_tokens": 63625591.0,
      "step": 36675
    },
    {
      "entropy": 5.882905912399292,
      "epoch": 2.853841665045711,
      "grad_norm": 1.25,
      "learning_rate": 0.0004185339766990794,
      "loss": 5.137,
      "mean_token_accuracy": 0.1969687297940254,
      "num_tokens": 63633681.0,
      "step": 36680
    },
    {
      "entropy": 5.793805408477783,
      "epoch": 2.8542306944174287,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004185126319647213,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20480012148618698,
      "num_tokens": 63642298.0,
      "step": 36685
    },
    {
      "entropy": 5.822042179107666,
      "epoch": 2.854619723789146,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004184912850527875,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1998739406466484,
      "num_tokens": 63650969.0,
      "step": 36690
    },
    {
      "entropy": 5.742236948013305,
      "epoch": 2.8550087531608637,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004184699359636018,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21801754385232924,
      "num_tokens": 63659477.0,
      "step": 36695
    },
    {
      "entropy": 5.7641764163970945,
      "epoch": 2.8553977825325814,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004184485846974882,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20335123240947722,
      "num_tokens": 63668575.0,
      "step": 36700
    },
    {
      "entropy": 5.796372699737549,
      "epoch": 2.8557868119042986,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004184272312547706,
      "loss": 5.0739,
      "mean_token_accuracy": 0.20363492518663406,
      "num_tokens": 63677818.0,
      "step": 36705
    },
    {
      "entropy": 5.785946702957153,
      "epoch": 2.8561758412760163,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041840587563577315,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21071430146694184,
      "num_tokens": 63686360.0,
      "step": 36710
    },
    {
      "entropy": 5.748115110397339,
      "epoch": 2.856564870647734,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004183845178408197,
      "loss": 5.0643,
      "mean_token_accuracy": 0.19669489860534667,
      "num_tokens": 63695602.0,
      "step": 36715
    },
    {
      "entropy": 5.747306632995605,
      "epoch": 2.8569539000194517,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041836315787023456,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21039898544549943,
      "num_tokens": 63704225.0,
      "step": 36720
    },
    {
      "entropy": 5.784542608261108,
      "epoch": 2.857342929391169,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041834179572434153,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2053074210882187,
      "num_tokens": 63712814.0,
      "step": 36725
    },
    {
      "entropy": 5.768649530410767,
      "epoch": 2.8577319587628867,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041832043140346495,
      "loss": 4.9604,
      "mean_token_accuracy": 0.20785662531852722,
      "num_tokens": 63721808.0,
      "step": 36730
    },
    {
      "entropy": 5.7266960620880125,
      "epoch": 2.858120988134604,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004182990649079289,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20898027569055558,
      "num_tokens": 63730226.0,
      "step": 36735
    },
    {
      "entropy": 5.739819478988648,
      "epoch": 2.8585100175063216,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004182776962380577,
      "loss": 4.9443,
      "mean_token_accuracy": 0.211281980574131,
      "num_tokens": 63738400.0,
      "step": 36740
    },
    {
      "entropy": 5.714207696914673,
      "epoch": 2.8588990468780393,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041825632539417546,
      "loss": 4.9546,
      "mean_token_accuracy": 0.2127688303589821,
      "num_tokens": 63746385.0,
      "step": 36745
    },
    {
      "entropy": 5.771510362625122,
      "epoch": 2.859288076249757,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004182349523766065,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20162491500377655,
      "num_tokens": 63755354.0,
      "step": 36750
    },
    {
      "entropy": 5.865142440795898,
      "epoch": 2.8596771056214743,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041821357718567514,
      "loss": 5.0905,
      "mean_token_accuracy": 0.2023327976465225,
      "num_tokens": 63764262.0,
      "step": 36755
    },
    {
      "entropy": 5.762549257278442,
      "epoch": 2.860066134993192,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004181921998217057,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21103435754776,
      "num_tokens": 63772770.0,
      "step": 36760
    },
    {
      "entropy": 5.758109188079834,
      "epoch": 2.8604551643649097,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004181708202850225,
      "loss": 4.9893,
      "mean_token_accuracy": 0.21005165278911592,
      "num_tokens": 63781353.0,
      "step": 36765
    },
    {
      "entropy": 5.859723854064941,
      "epoch": 2.860844193736627,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041814943857595,
      "loss": 5.0984,
      "mean_token_accuracy": 0.2010166496038437,
      "num_tokens": 63789262.0,
      "step": 36770
    },
    {
      "entropy": 5.866224145889282,
      "epoch": 2.8612332231083446,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004181280546948127,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1964018926024437,
      "num_tokens": 63797792.0,
      "step": 36775
    },
    {
      "entropy": 5.7840923309326175,
      "epoch": 2.8616222524800623,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004181066686419349,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2104126051068306,
      "num_tokens": 63805362.0,
      "step": 36780
    },
    {
      "entropy": 5.762281274795532,
      "epoch": 2.86201128185178,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041808528041764115,
      "loss": 5.014,
      "mean_token_accuracy": 0.20400736778974532,
      "num_tokens": 63813942.0,
      "step": 36785
    },
    {
      "entropy": 5.657808113098144,
      "epoch": 2.8624003112234973,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004180638900222561,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22111211121082305,
      "num_tokens": 63823077.0,
      "step": 36790
    },
    {
      "entropy": 5.683550071716309,
      "epoch": 2.862789340595215,
      "grad_norm": 1.25,
      "learning_rate": 0.0004180424974561042,
      "loss": 4.9015,
      "mean_token_accuracy": 0.21220102906227112,
      "num_tokens": 63831048.0,
      "step": 36795
    },
    {
      "entropy": 5.783267402648926,
      "epoch": 2.8631783699669326,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041802110271951,
      "loss": 5.0388,
      "mean_token_accuracy": 0.19883078038692475,
      "num_tokens": 63839449.0,
      "step": 36800
    },
    {
      "entropy": 5.7429337978363035,
      "epoch": 2.86356739933865,
      "grad_norm": 1.375,
      "learning_rate": 0.0004179997058127983,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2107088163495064,
      "num_tokens": 63847945.0,
      "step": 36805
    },
    {
      "entropy": 5.714373636245727,
      "epoch": 2.8639564287103676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004179783067362936,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21662425398826599,
      "num_tokens": 63857334.0,
      "step": 36810
    },
    {
      "entropy": 5.7682861328125,
      "epoch": 2.8643454580820853,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004179569054903207,
      "loss": 5.046,
      "mean_token_accuracy": 0.19507195800542831,
      "num_tokens": 63866272.0,
      "step": 36815
    },
    {
      "entropy": 5.759061098098755,
      "epoch": 2.864734487453803,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041793550207520436,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21360572874546052,
      "num_tokens": 63874798.0,
      "step": 36820
    },
    {
      "entropy": 5.776929903030395,
      "epoch": 2.8651235168255202,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004179140964912692,
      "loss": 4.958,
      "mean_token_accuracy": 0.21114621311426163,
      "num_tokens": 63883106.0,
      "step": 36825
    },
    {
      "entropy": 5.772731304168701,
      "epoch": 2.865512546197238,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004178926887388401,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21118485033512116,
      "num_tokens": 63892172.0,
      "step": 36830
    },
    {
      "entropy": 5.8111998558044435,
      "epoch": 2.865901575568955,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00041787127881824193,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20013560056686402,
      "num_tokens": 63900436.0,
      "step": 36835
    },
    {
      "entropy": 5.826224946975708,
      "epoch": 2.866290604940673,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041784986672979945,
      "loss": 5.0657,
      "mean_token_accuracy": 0.201530821621418,
      "num_tokens": 63908937.0,
      "step": 36840
    },
    {
      "entropy": 5.757696962356567,
      "epoch": 2.8666796343123906,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041782845247383767,
      "loss": 5.03,
      "mean_token_accuracy": 0.2060224696993828,
      "num_tokens": 63916350.0,
      "step": 36845
    },
    {
      "entropy": 5.752378511428833,
      "epoch": 2.8670686636841083,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041780703605068145,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20287578105926513,
      "num_tokens": 63925227.0,
      "step": 36850
    },
    {
      "entropy": 5.8021773338317875,
      "epoch": 2.867457693055826,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004177856174606558,
      "loss": 4.993,
      "mean_token_accuracy": 0.20400072485208512,
      "num_tokens": 63934075.0,
      "step": 36855
    },
    {
      "entropy": 5.757309150695801,
      "epoch": 2.8678467224275432,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004177641967040856,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20339630991220475,
      "num_tokens": 63942554.0,
      "step": 36860
    },
    {
      "entropy": 5.720445823669434,
      "epoch": 2.868235751799261,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041774277378129595,
      "loss": 4.933,
      "mean_token_accuracy": 0.2095974788069725,
      "num_tokens": 63951850.0,
      "step": 36865
    },
    {
      "entropy": 5.817222690582275,
      "epoch": 2.868624781170978,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041772134869261186,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2073666647076607,
      "num_tokens": 63960709.0,
      "step": 36870
    },
    {
      "entropy": 5.757843542098999,
      "epoch": 2.869013810542696,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004176999214383585,
      "loss": 4.9678,
      "mean_token_accuracy": 0.2075099006295204,
      "num_tokens": 63969859.0,
      "step": 36875
    },
    {
      "entropy": 5.765377187728882,
      "epoch": 2.8694028399144136,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041767849201886103,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20766817927360534,
      "num_tokens": 63977524.0,
      "step": 36880
    },
    {
      "entropy": 5.666045427322388,
      "epoch": 2.8697918692861313,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004176570604344445,
      "loss": 4.8629,
      "mean_token_accuracy": 0.21466914117336272,
      "num_tokens": 63985938.0,
      "step": 36885
    },
    {
      "entropy": 5.684049940109253,
      "epoch": 2.8701808986578485,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041763562668543415,
      "loss": 5.0189,
      "mean_token_accuracy": 0.20620413720607758,
      "num_tokens": 63993864.0,
      "step": 36890
    },
    {
      "entropy": 5.801980590820312,
      "epoch": 2.8705699280295662,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004176141907721552,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20229703933000565,
      "num_tokens": 64003479.0,
      "step": 36895
    },
    {
      "entropy": 5.840611934661865,
      "epoch": 2.870958957401284,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004175927526949329,
      "loss": 4.995,
      "mean_token_accuracy": 0.2094767287373543,
      "num_tokens": 64012492.0,
      "step": 36900
    },
    {
      "entropy": 5.827215480804443,
      "epoch": 2.871347986773001,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004175713124540925,
      "loss": 5.0981,
      "mean_token_accuracy": 0.19082699120044708,
      "num_tokens": 64021756.0,
      "step": 36905
    },
    {
      "entropy": 5.791811180114746,
      "epoch": 2.871737016144719,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041754987004995935,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20265102088451387,
      "num_tokens": 64030351.0,
      "step": 36910
    },
    {
      "entropy": 5.734309530258178,
      "epoch": 2.8721260455164366,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041752842548285887,
      "loss": 4.9176,
      "mean_token_accuracy": 0.20751064717769624,
      "num_tokens": 64039523.0,
      "step": 36915
    },
    {
      "entropy": 5.682511949539185,
      "epoch": 2.8725150748881543,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004175069787531163,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2154543086886406,
      "num_tokens": 64048264.0,
      "step": 36920
    },
    {
      "entropy": 5.742174434661865,
      "epoch": 2.8729041042598715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041748552986105726,
      "loss": 5.0518,
      "mean_token_accuracy": 0.19745032340288163,
      "num_tokens": 64057013.0,
      "step": 36925
    },
    {
      "entropy": 5.7233950138092045,
      "epoch": 2.8732931336315892,
      "grad_norm": 1.296875,
      "learning_rate": 0.000417464078807007,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21013311445713043,
      "num_tokens": 64065962.0,
      "step": 36930
    },
    {
      "entropy": 5.749708557128907,
      "epoch": 2.8736821630033065,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004174426255912912,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21255343705415725,
      "num_tokens": 64074493.0,
      "step": 36935
    },
    {
      "entropy": 5.813030147552491,
      "epoch": 2.874071192375024,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004174211702142352,
      "loss": 5.1002,
      "mean_token_accuracy": 0.19952668994665146,
      "num_tokens": 64083351.0,
      "step": 36940
    },
    {
      "entropy": 5.835479402542115,
      "epoch": 2.874460221746742,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041739971267616466,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21079067885875702,
      "num_tokens": 64091775.0,
      "step": 36945
    },
    {
      "entropy": 5.669658374786377,
      "epoch": 2.8748492511184596,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004173782529774051,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2111445516347885,
      "num_tokens": 64100684.0,
      "step": 36950
    },
    {
      "entropy": 5.670276927947998,
      "epoch": 2.8752382804901773,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041735679111828223,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21021228581666945,
      "num_tokens": 64109102.0,
      "step": 36955
    },
    {
      "entropy": 5.703567743301392,
      "epoch": 2.8756273098618945,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041733532709912157,
      "loss": 4.9693,
      "mean_token_accuracy": 0.20582934468984604,
      "num_tokens": 64118341.0,
      "step": 36960
    },
    {
      "entropy": 5.7525248527526855,
      "epoch": 2.876016339233612,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041731386092024893,
      "loss": 4.8806,
      "mean_token_accuracy": 0.21486097127199172,
      "num_tokens": 64126647.0,
      "step": 36965
    },
    {
      "entropy": 5.657056474685669,
      "epoch": 2.8764053686053295,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041729239258198996,
      "loss": 4.8139,
      "mean_token_accuracy": 0.21374335438013076,
      "num_tokens": 64135534.0,
      "step": 36970
    },
    {
      "entropy": 5.7172324657440186,
      "epoch": 2.876794397977047,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041727092208467036,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20769294500350952,
      "num_tokens": 64144108.0,
      "step": 36975
    },
    {
      "entropy": 5.706579971313476,
      "epoch": 2.877183427348765,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041724944942861603,
      "loss": 4.8523,
      "mean_token_accuracy": 0.21420302540063857,
      "num_tokens": 64152599.0,
      "step": 36980
    },
    {
      "entropy": 5.753954601287842,
      "epoch": 2.8775724567204826,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041722797461415267,
      "loss": 4.9821,
      "mean_token_accuracy": 0.2031369611620903,
      "num_tokens": 64161733.0,
      "step": 36985
    },
    {
      "entropy": 5.758615398406983,
      "epoch": 2.8779614860922,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004172064976416062,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21153523176908492,
      "num_tokens": 64171045.0,
      "step": 36990
    },
    {
      "entropy": 5.777635288238526,
      "epoch": 2.8783505154639175,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004171850185113024,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21195402145385742,
      "num_tokens": 64179725.0,
      "step": 36995
    },
    {
      "entropy": 5.728305149078369,
      "epoch": 2.878739544835635,
      "grad_norm": 1.25,
      "learning_rate": 0.00041716353722356735,
      "loss": 5.0088,
      "mean_token_accuracy": 0.20766429752111434,
      "num_tokens": 64188799.0,
      "step": 37000
    },
    {
      "entropy": 5.791095781326294,
      "epoch": 2.8791285742073525,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004171420537787269,
      "loss": 5.0403,
      "mean_token_accuracy": 0.20113651752471923,
      "num_tokens": 64197809.0,
      "step": 37005
    },
    {
      "entropy": 5.7563423156738285,
      "epoch": 2.87951760357907,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041712056817710697,
      "loss": 5.0597,
      "mean_token_accuracy": 0.19558840543031691,
      "num_tokens": 64207113.0,
      "step": 37010
    },
    {
      "entropy": 5.802539539337158,
      "epoch": 2.879906632950788,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004170990804190337,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20976713597774505,
      "num_tokens": 64216846.0,
      "step": 37015
    },
    {
      "entropy": 5.753994035720825,
      "epoch": 2.8802956623225056,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041707759050483304,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20872389823198317,
      "num_tokens": 64225348.0,
      "step": 37020
    },
    {
      "entropy": 5.739677429199219,
      "epoch": 2.880684691694223,
      "grad_norm": 1.296875,
      "learning_rate": 0.000417056098434831,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20938349962234498,
      "num_tokens": 64234065.0,
      "step": 37025
    },
    {
      "entropy": 5.754743242263794,
      "epoch": 2.8810737210659405,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004170346042093538,
      "loss": 4.9914,
      "mean_token_accuracy": 0.20214109718799592,
      "num_tokens": 64242484.0,
      "step": 37030
    },
    {
      "entropy": 5.691471338272095,
      "epoch": 2.881462750437658,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004170131078287276,
      "loss": 4.8713,
      "mean_token_accuracy": 0.21622071862220765,
      "num_tokens": 64251056.0,
      "step": 37035
    },
    {
      "entropy": 5.771268701553344,
      "epoch": 2.8818517798093755,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004169916092932785,
      "loss": 4.955,
      "mean_token_accuracy": 0.2091274991631508,
      "num_tokens": 64258961.0,
      "step": 37040
    },
    {
      "entropy": 5.6980125427246096,
      "epoch": 2.882240809181093,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004169701086033327,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21173637360334396,
      "num_tokens": 64268133.0,
      "step": 37045
    },
    {
      "entropy": 5.796798658370972,
      "epoch": 2.882629838552811,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041694860575921654,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20618970990180968,
      "num_tokens": 64276487.0,
      "step": 37050
    },
    {
      "entropy": 5.7684250831604,
      "epoch": 2.8830188679245285,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041692710076125615,
      "loss": 4.9422,
      "mean_token_accuracy": 0.21119063794612886,
      "num_tokens": 64284983.0,
      "step": 37055
    },
    {
      "entropy": 5.754338169097901,
      "epoch": 2.883407897296246,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000416905593609778,
      "loss": 5.0158,
      "mean_token_accuracy": 0.20715288519859315,
      "num_tokens": 64293875.0,
      "step": 37060
    },
    {
      "entropy": 5.786267709732056,
      "epoch": 2.8837969266679635,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041688408430510827,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21430833339691163,
      "num_tokens": 64302628.0,
      "step": 37065
    },
    {
      "entropy": 5.781233167648315,
      "epoch": 2.8841859560396808,
      "grad_norm": 1.25,
      "learning_rate": 0.0004168625728475734,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2082262307405472,
      "num_tokens": 64311024.0,
      "step": 37070
    },
    {
      "entropy": 5.721941947937012,
      "epoch": 2.8845749854113985,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041684105923749977,
      "loss": 4.9946,
      "mean_token_accuracy": 0.20572635233402253,
      "num_tokens": 64318836.0,
      "step": 37075
    },
    {
      "entropy": 5.729269552230835,
      "epoch": 2.884964014783116,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004168195434752139,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21010118573904038,
      "num_tokens": 64326983.0,
      "step": 37080
    },
    {
      "entropy": 5.782969284057617,
      "epoch": 2.885353044154834,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041679802556104207,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20863622277975083,
      "num_tokens": 64335304.0,
      "step": 37085
    },
    {
      "entropy": 5.782275676727295,
      "epoch": 2.885742073526551,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041677650549531097,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21518247127532958,
      "num_tokens": 64343515.0,
      "step": 37090
    },
    {
      "entropy": 5.7294168949127195,
      "epoch": 2.886131102898269,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041675498327834706,
      "loss": 4.9919,
      "mean_token_accuracy": 0.20612918883562087,
      "num_tokens": 64351936.0,
      "step": 37095
    },
    {
      "entropy": 5.809228563308716,
      "epoch": 2.8865201322699865,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041673345891047686,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20429991483688353,
      "num_tokens": 64360332.0,
      "step": 37100
    },
    {
      "entropy": 5.804297161102295,
      "epoch": 2.8869091616417037,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004167119323920271,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20966865569353105,
      "num_tokens": 64369615.0,
      "step": 37105
    },
    {
      "entropy": 5.743694114685058,
      "epoch": 2.8872981910134214,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004166904037233243,
      "loss": 4.9251,
      "mean_token_accuracy": 0.20841394513845443,
      "num_tokens": 64378355.0,
      "step": 37110
    },
    {
      "entropy": 5.712008047103882,
      "epoch": 2.887687220385139,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041666887290469506,
      "loss": 4.9049,
      "mean_token_accuracy": 0.21058732718229295,
      "num_tokens": 64386120.0,
      "step": 37115
    },
    {
      "entropy": 5.6964539051055905,
      "epoch": 2.888076249756857,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004166473399364663,
      "loss": 4.9829,
      "mean_token_accuracy": 0.20607653707265855,
      "num_tokens": 64395153.0,
      "step": 37120
    },
    {
      "entropy": 5.810642528533935,
      "epoch": 2.888465279128574,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041662580481896446,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20461305528879165,
      "num_tokens": 64404266.0,
      "step": 37125
    },
    {
      "entropy": 5.771422481536865,
      "epoch": 2.888854308500292,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004166042675525165,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20623352080583573,
      "num_tokens": 64412999.0,
      "step": 37130
    },
    {
      "entropy": 5.770464038848877,
      "epoch": 2.8892433378720095,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041658272813744924,
      "loss": 4.9823,
      "mean_token_accuracy": 0.20960859060287476,
      "num_tokens": 64421561.0,
      "step": 37135
    },
    {
      "entropy": 5.749947595596313,
      "epoch": 2.8896323672437267,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004165611865740894,
      "loss": 5.0388,
      "mean_token_accuracy": 0.20452422350645066,
      "num_tokens": 64430254.0,
      "step": 37140
    },
    {
      "entropy": 5.760941505432129,
      "epoch": 2.8900213966154444,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004165396428627638,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21514674425125122,
      "num_tokens": 64438452.0,
      "step": 37145
    },
    {
      "entropy": 5.867004013061523,
      "epoch": 2.890410425987162,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041651809700379947,
      "loss": 5.0978,
      "mean_token_accuracy": 0.20511178374290467,
      "num_tokens": 64447442.0,
      "step": 37150
    },
    {
      "entropy": 5.7775932312011715,
      "epoch": 2.89079945535888,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041649654899752326,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20124804973602295,
      "num_tokens": 64456179.0,
      "step": 37155
    },
    {
      "entropy": 5.871944713592529,
      "epoch": 2.891188484730597,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004164749988442621,
      "loss": 5.1187,
      "mean_token_accuracy": 0.19993913620710374,
      "num_tokens": 64464498.0,
      "step": 37160
    },
    {
      "entropy": 5.760349178314209,
      "epoch": 2.891577514102315,
      "grad_norm": 1.3125,
      "learning_rate": 0.000416453446544343,
      "loss": 4.9577,
      "mean_token_accuracy": 0.19800868928432463,
      "num_tokens": 64473047.0,
      "step": 37165
    },
    {
      "entropy": 5.680169248580933,
      "epoch": 2.891966543474032,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004164318920980931,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22016491293907164,
      "num_tokens": 64482338.0,
      "step": 37170
    },
    {
      "entropy": 5.7133036136627195,
      "epoch": 2.8923555728457497,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004164103355058393,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2117451786994934,
      "num_tokens": 64490611.0,
      "step": 37175
    },
    {
      "entropy": 5.711759281158447,
      "epoch": 2.8927446022174674,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004163887767679087,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2107089191675186,
      "num_tokens": 64499180.0,
      "step": 37180
    },
    {
      "entropy": 5.798609638214112,
      "epoch": 2.893133631589185,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004163672158846285,
      "loss": 5.0194,
      "mean_token_accuracy": 0.19753916412591935,
      "num_tokens": 64508664.0,
      "step": 37185
    },
    {
      "entropy": 5.91753044128418,
      "epoch": 2.8935226609609024,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00041634565285632573,
      "loss": 5.125,
      "mean_token_accuracy": 0.19866318255662918,
      "num_tokens": 64518059.0,
      "step": 37190
    },
    {
      "entropy": 5.820940732955933,
      "epoch": 2.89391169033262,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041632408768332776,
      "loss": 5.0985,
      "mean_token_accuracy": 0.1998754099011421,
      "num_tokens": 64526616.0,
      "step": 37195
    },
    {
      "entropy": 5.854300212860108,
      "epoch": 2.8943007197043378,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041630252036596165,
      "loss": 5.0818,
      "mean_token_accuracy": 0.20094608068466185,
      "num_tokens": 64535008.0,
      "step": 37200
    },
    {
      "entropy": 5.759271478652954,
      "epoch": 2.894689749076055,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004162809509045547,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21034590750932694,
      "num_tokens": 64543741.0,
      "step": 37205
    },
    {
      "entropy": 5.673016977310181,
      "epoch": 2.8950787784477727,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041625937929943424,
      "loss": 4.8451,
      "mean_token_accuracy": 0.21106300950050355,
      "num_tokens": 64552409.0,
      "step": 37210
    },
    {
      "entropy": 5.781196022033692,
      "epoch": 2.8954678078194904,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041623780555092747,
      "loss": 5.0796,
      "mean_token_accuracy": 0.20001500695943833,
      "num_tokens": 64561308.0,
      "step": 37215
    },
    {
      "entropy": 5.761783075332642,
      "epoch": 2.895856837191208,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004162162296593618,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21576246172189711,
      "num_tokens": 64569671.0,
      "step": 37220
    },
    {
      "entropy": 5.867943477630615,
      "epoch": 2.8962458665629254,
      "grad_norm": 1.25,
      "learning_rate": 0.00041619465162506466,
      "loss": 5.0694,
      "mean_token_accuracy": 0.2004273295402527,
      "num_tokens": 64578186.0,
      "step": 37225
    },
    {
      "entropy": 5.7739870071411135,
      "epoch": 2.896634895934643,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004161730714483633,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21497567594051362,
      "num_tokens": 64586703.0,
      "step": 37230
    },
    {
      "entropy": 5.71979832649231,
      "epoch": 2.8970239253063608,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004161514891295854,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20912373960018157,
      "num_tokens": 64595119.0,
      "step": 37235
    },
    {
      "entropy": 5.794033241271973,
      "epoch": 2.897412954678078,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041612990466905825,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20521072149276734,
      "num_tokens": 64604312.0,
      "step": 37240
    },
    {
      "entropy": 5.805680131912231,
      "epoch": 2.8978019840497957,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041610831806710954,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2112719863653183,
      "num_tokens": 64613088.0,
      "step": 37245
    },
    {
      "entropy": 5.863837432861328,
      "epoch": 2.8981910134215134,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041608672932406666,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1959698721766472,
      "num_tokens": 64621659.0,
      "step": 37250
    },
    {
      "entropy": 5.77723536491394,
      "epoch": 2.898580042793231,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041606513844025716,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21353244185447692,
      "num_tokens": 64630436.0,
      "step": 37255
    },
    {
      "entropy": 5.723501586914063,
      "epoch": 2.8989690721649484,
      "grad_norm": 1.25,
      "learning_rate": 0.0004160435454160087,
      "loss": 4.99,
      "mean_token_accuracy": 0.21496034562587737,
      "num_tokens": 64639836.0,
      "step": 37260
    },
    {
      "entropy": 5.763289928436279,
      "epoch": 2.899358101536666,
      "grad_norm": 1.203125,
      "learning_rate": 0.000416021950251649,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2007473587989807,
      "num_tokens": 64648941.0,
      "step": 37265
    },
    {
      "entropy": 5.74602370262146,
      "epoch": 2.8997471309083833,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004160003529475057,
      "loss": 5.039,
      "mean_token_accuracy": 0.1992994502186775,
      "num_tokens": 64657632.0,
      "step": 37270
    },
    {
      "entropy": 5.750827407836914,
      "epoch": 2.900136160280101,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004159787535039064,
      "loss": 5.023,
      "mean_token_accuracy": 0.19981631636619568,
      "num_tokens": 64665733.0,
      "step": 37275
    },
    {
      "entropy": 5.637521028518677,
      "epoch": 2.9005251896518187,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004159571519211789,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2171190321445465,
      "num_tokens": 64673873.0,
      "step": 37280
    },
    {
      "entropy": 5.760395336151123,
      "epoch": 2.9009142190235364,
      "grad_norm": 1.296875,
      "learning_rate": 0.000415935548199651,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20852830708026887,
      "num_tokens": 64682282.0,
      "step": 37285
    },
    {
      "entropy": 5.731938695907592,
      "epoch": 2.901303248395254,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004159139423396504,
      "loss": 4.9439,
      "mean_token_accuracy": 0.20763493925333024,
      "num_tokens": 64690179.0,
      "step": 37290
    },
    {
      "entropy": 5.76706748008728,
      "epoch": 2.9016922777669714,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004158923343415051,
      "loss": 5.075,
      "mean_token_accuracy": 0.20487562119960784,
      "num_tokens": 64698935.0,
      "step": 37295
    },
    {
      "entropy": 5.699236345291138,
      "epoch": 2.902081307138689,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004158707242055429,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21039014756679536,
      "num_tokens": 64708328.0,
      "step": 37300
    },
    {
      "entropy": 5.822888517379761,
      "epoch": 2.9024703365104063,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041584911193209167,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2005120113492012,
      "num_tokens": 64717003.0,
      "step": 37305
    },
    {
      "entropy": 5.778169775009156,
      "epoch": 2.902859365882124,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004158274975214794,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20894182920455934,
      "num_tokens": 64725951.0,
      "step": 37310
    },
    {
      "entropy": 5.771015787124634,
      "epoch": 2.9032483952538417,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041580588097403384,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20769935846328735,
      "num_tokens": 64733949.0,
      "step": 37315
    },
    {
      "entropy": 5.775215148925781,
      "epoch": 2.9036374246255594,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041578426229008325,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20621199756860734,
      "num_tokens": 64742729.0,
      "step": 37320
    },
    {
      "entropy": 5.782303142547607,
      "epoch": 2.9040264539972767,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041576264146995557,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2029201328754425,
      "num_tokens": 64751655.0,
      "step": 37325
    },
    {
      "entropy": 5.817810344696045,
      "epoch": 2.9044154833689944,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041574101851397883,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2116365283727646,
      "num_tokens": 64760019.0,
      "step": 37330
    },
    {
      "entropy": 5.766209077835083,
      "epoch": 2.904804512740712,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004157193934224811,
      "loss": 5.0209,
      "mean_token_accuracy": 0.21017147302627565,
      "num_tokens": 64768528.0,
      "step": 37335
    },
    {
      "entropy": 5.789718198776245,
      "epoch": 2.9051935421124293,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004156977661957906,
      "loss": 5.0179,
      "mean_token_accuracy": 0.20197457820177078,
      "num_tokens": 64776903.0,
      "step": 37340
    },
    {
      "entropy": 5.831532430648804,
      "epoch": 2.905582571484147,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004156761368342355,
      "loss": 5.1278,
      "mean_token_accuracy": 0.19620117396116257,
      "num_tokens": 64785361.0,
      "step": 37345
    },
    {
      "entropy": 5.8084814071655275,
      "epoch": 2.9059716008558647,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041565450533814384,
      "loss": 5.0473,
      "mean_token_accuracy": 0.203514988720417,
      "num_tokens": 64793720.0,
      "step": 37350
    },
    {
      "entropy": 5.72900128364563,
      "epoch": 2.9063606302275824,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041563287170784396,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2146420881152153,
      "num_tokens": 64802049.0,
      "step": 37355
    },
    {
      "entropy": 5.755378770828247,
      "epoch": 2.9067496595992997,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004156112359436641,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20473323613405228,
      "num_tokens": 64810754.0,
      "step": 37360
    },
    {
      "entropy": 5.825936555862427,
      "epoch": 2.9071386889710173,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041558959804593255,
      "loss": 5.0551,
      "mean_token_accuracy": 0.19938816130161285,
      "num_tokens": 64819466.0,
      "step": 37365
    },
    {
      "entropy": 5.788001251220703,
      "epoch": 2.9075277183427346,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00041556795801497766,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2044745296239853,
      "num_tokens": 64828135.0,
      "step": 37370
    },
    {
      "entropy": 5.797698450088501,
      "epoch": 2.9079167477144523,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004155463158511277,
      "loss": 5.0716,
      "mean_token_accuracy": 0.20172516256570816,
      "num_tokens": 64836542.0,
      "step": 37375
    },
    {
      "entropy": 5.754073429107666,
      "epoch": 2.90830577708617,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041552467155471105,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21227678805589675,
      "num_tokens": 64844470.0,
      "step": 37380
    },
    {
      "entropy": 5.748687839508056,
      "epoch": 2.9086948064578877,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004155030251260563,
      "loss": 4.994,
      "mean_token_accuracy": 0.20368156880140303,
      "num_tokens": 64854011.0,
      "step": 37385
    },
    {
      "entropy": 5.830940008163452,
      "epoch": 2.9090838358296054,
      "grad_norm": 1.25,
      "learning_rate": 0.00041548137656549167,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2085736483335495,
      "num_tokens": 64862349.0,
      "step": 37390
    },
    {
      "entropy": 5.826216125488282,
      "epoch": 2.9094728652013226,
      "grad_norm": 1.375,
      "learning_rate": 0.0004154597258733457,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20644983947277068,
      "num_tokens": 64870634.0,
      "step": 37395
    },
    {
      "entropy": 5.807198905944825,
      "epoch": 2.9098618945730403,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004154380730499472,
      "loss": 5.0548,
      "mean_token_accuracy": 0.20213259309530257,
      "num_tokens": 64879870.0,
      "step": 37400
    },
    {
      "entropy": 5.729497337341309,
      "epoch": 2.9102509239447576,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041541641809562425,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21021427363157272,
      "num_tokens": 64888650.0,
      "step": 37405
    },
    {
      "entropy": 5.761759519577026,
      "epoch": 2.9106399533164753,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004153947610107057,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21261070519685746,
      "num_tokens": 64896788.0,
      "step": 37410
    },
    {
      "entropy": 5.739074802398681,
      "epoch": 2.911028982688193,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041537310179552017,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22130744457244872,
      "num_tokens": 64905134.0,
      "step": 37415
    },
    {
      "entropy": 5.7758180618286135,
      "epoch": 2.9114180120599107,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004153514404503962,
      "loss": 5.0467,
      "mean_token_accuracy": 0.2032207116484642,
      "num_tokens": 64913662.0,
      "step": 37420
    },
    {
      "entropy": 5.756581020355225,
      "epoch": 2.911807041431628,
      "grad_norm": 1.25,
      "learning_rate": 0.0004153297769756625,
      "loss": 4.9632,
      "mean_token_accuracy": 0.21150299310684204,
      "num_tokens": 64922159.0,
      "step": 37425
    },
    {
      "entropy": 5.7621971607208256,
      "epoch": 2.9121960708033456,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004153081113716478,
      "loss": 4.932,
      "mean_token_accuracy": 0.213905930519104,
      "num_tokens": 64930672.0,
      "step": 37430
    },
    {
      "entropy": 5.74616813659668,
      "epoch": 2.9125851001750633,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041528644363868087,
      "loss": 4.9926,
      "mean_token_accuracy": 0.20725370645523072,
      "num_tokens": 64940424.0,
      "step": 37435
    },
    {
      "entropy": 5.813746547698974,
      "epoch": 2.9129741295467806,
      "grad_norm": 1.25,
      "learning_rate": 0.0004152647737770904,
      "loss": 5.1229,
      "mean_token_accuracy": 0.20157860219478607,
      "num_tokens": 64949510.0,
      "step": 37440
    },
    {
      "entropy": 5.71826400756836,
      "epoch": 2.9133631589184983,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004152431017872053,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21279249787330629,
      "num_tokens": 64958240.0,
      "step": 37445
    },
    {
      "entropy": 5.7577433586120605,
      "epoch": 2.913752188290216,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041522142766935426,
      "loss": 4.9521,
      "mean_token_accuracy": 0.20746141225099562,
      "num_tokens": 64967071.0,
      "step": 37450
    },
    {
      "entropy": 5.651897382736206,
      "epoch": 2.9141412176619337,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041519975142386627,
      "loss": 4.966,
      "mean_token_accuracy": 0.21062342524528505,
      "num_tokens": 64974949.0,
      "step": 37455
    },
    {
      "entropy": 5.806445026397705,
      "epoch": 2.914530247033651,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004151780730510702,
      "loss": 5.0159,
      "mean_token_accuracy": 0.19842678308486938,
      "num_tokens": 64983510.0,
      "step": 37460
    },
    {
      "entropy": 5.8871102809906,
      "epoch": 2.9149192764053686,
      "grad_norm": 1.25,
      "learning_rate": 0.00041515639255129504,
      "loss": 5.1122,
      "mean_token_accuracy": 0.20638594776391983,
      "num_tokens": 64993055.0,
      "step": 37465
    },
    {
      "entropy": 5.823628664016724,
      "epoch": 2.9153083057770863,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004151347099248697,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20884522795677185,
      "num_tokens": 65002406.0,
      "step": 37470
    },
    {
      "entropy": 5.784718084335327,
      "epoch": 2.9156973351488036,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004151130251721231,
      "loss": 5.1003,
      "mean_token_accuracy": 0.19495654553174974,
      "num_tokens": 65011949.0,
      "step": 37475
    },
    {
      "entropy": 5.8989917755126955,
      "epoch": 2.9160863645205213,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041509133829338454,
      "loss": 5.1265,
      "mean_token_accuracy": 0.1953040137887001,
      "num_tokens": 65020215.0,
      "step": 37480
    },
    {
      "entropy": 5.788927268981934,
      "epoch": 2.916475393892239,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004150696492889827,
      "loss": 4.9702,
      "mean_token_accuracy": 0.21589181423187256,
      "num_tokens": 65028423.0,
      "step": 37485
    },
    {
      "entropy": 5.6992881298065186,
      "epoch": 2.9168644232639567,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041504795815924697,
      "loss": 4.9418,
      "mean_token_accuracy": 0.2089121699333191,
      "num_tokens": 65037804.0,
      "step": 37490
    },
    {
      "entropy": 5.727133941650391,
      "epoch": 2.917253452635674,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041502626490450635,
      "loss": 4.9104,
      "mean_token_accuracy": 0.21570639163255692,
      "num_tokens": 65046389.0,
      "step": 37495
    },
    {
      "entropy": 5.733090591430664,
      "epoch": 2.9176424820073916,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004150045695250901,
      "loss": 4.9458,
      "mean_token_accuracy": 0.20948667377233504,
      "num_tokens": 65055499.0,
      "step": 37500
    },
    {
      "entropy": 5.740337276458741,
      "epoch": 2.918031511379109,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004149828720213272,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21627739071846008,
      "num_tokens": 65063170.0,
      "step": 37505
    },
    {
      "entropy": 5.762694263458252,
      "epoch": 2.9184205407508266,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004149611723935472,
      "loss": 5.0314,
      "mean_token_accuracy": 0.20729401111602783,
      "num_tokens": 65071046.0,
      "step": 37510
    },
    {
      "entropy": 5.685011768341065,
      "epoch": 2.9188095701225443,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004149394706420791,
      "loss": 4.9919,
      "mean_token_accuracy": 0.20930340141057968,
      "num_tokens": 65079135.0,
      "step": 37515
    },
    {
      "entropy": 5.77154049873352,
      "epoch": 2.919198599494262,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041491776676725223,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20170643776655198,
      "num_tokens": 65087861.0,
      "step": 37520
    },
    {
      "entropy": 5.77112979888916,
      "epoch": 2.9195876288659792,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004148960607693961,
      "loss": 4.936,
      "mean_token_accuracy": 0.21039224565029144,
      "num_tokens": 65096561.0,
      "step": 37525
    },
    {
      "entropy": 5.794776201248169,
      "epoch": 2.919976658237697,
      "grad_norm": 1.40625,
      "learning_rate": 0.00041487435264883974,
      "loss": 4.9876,
      "mean_token_accuracy": 0.2130448415875435,
      "num_tokens": 65105383.0,
      "step": 37530
    },
    {
      "entropy": 5.751872873306274,
      "epoch": 2.9203656876094146,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004148526424059128,
      "loss": 4.9586,
      "mean_token_accuracy": 0.20623200237751008,
      "num_tokens": 65113551.0,
      "step": 37535
    },
    {
      "entropy": 5.685251331329345,
      "epoch": 2.920754716981132,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004148309300409446,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21854791641235352,
      "num_tokens": 65121599.0,
      "step": 37540
    },
    {
      "entropy": 5.770877742767334,
      "epoch": 2.9211437463528496,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004148092155542646,
      "loss": 5.081,
      "mean_token_accuracy": 0.2036645606160164,
      "num_tokens": 65130881.0,
      "step": 37545
    },
    {
      "entropy": 5.743512916564941,
      "epoch": 2.9215327757245673,
      "grad_norm": 1.25,
      "learning_rate": 0.0004147874989462023,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20818814635276794,
      "num_tokens": 65139233.0,
      "step": 37550
    },
    {
      "entropy": 5.755264949798584,
      "epoch": 2.921921805096285,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004147657802170871,
      "loss": 4.9659,
      "mean_token_accuracy": 0.20452229976654052,
      "num_tokens": 65147738.0,
      "step": 37555
    },
    {
      "entropy": 5.79300537109375,
      "epoch": 2.922310834468002,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041474405936724873,
      "loss": 4.9859,
      "mean_token_accuracy": 0.20653869211673737,
      "num_tokens": 65156413.0,
      "step": 37560
    },
    {
      "entropy": 5.81798038482666,
      "epoch": 2.92269986383972,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004147223363970166,
      "loss": 5.0009,
      "mean_token_accuracy": 0.20680516362190246,
      "num_tokens": 65164614.0,
      "step": 37565
    },
    {
      "entropy": 5.794922399520874,
      "epoch": 2.9230888932114376,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004147006113067205,
      "loss": 5.0196,
      "mean_token_accuracy": 0.19798460453748704,
      "num_tokens": 65173159.0,
      "step": 37570
    },
    {
      "entropy": 5.711664915084839,
      "epoch": 2.923477922583155,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004146788840966901,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21148068010807036,
      "num_tokens": 65181999.0,
      "step": 37575
    },
    {
      "entropy": 5.84295163154602,
      "epoch": 2.9238669519548726,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004146571547672548,
      "loss": 5.155,
      "mean_token_accuracy": 0.18822210133075715,
      "num_tokens": 65191291.0,
      "step": 37580
    },
    {
      "entropy": 5.817547464370728,
      "epoch": 2.9242559813265903,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041463542331874443,
      "loss": 4.921,
      "mean_token_accuracy": 0.2133563593029976,
      "num_tokens": 65198891.0,
      "step": 37585
    },
    {
      "entropy": 5.764958190917969,
      "epoch": 2.924645010698308,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041461368975148876,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20982504338026048,
      "num_tokens": 65208206.0,
      "step": 37590
    },
    {
      "entropy": 5.718429088592529,
      "epoch": 2.925034040070025,
      "grad_norm": 1.375,
      "learning_rate": 0.00041459195406581763,
      "loss": 4.9561,
      "mean_token_accuracy": 0.20670429170131682,
      "num_tokens": 65216384.0,
      "step": 37595
    },
    {
      "entropy": 5.740876722335815,
      "epoch": 2.925423069441743,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004145702162620608,
      "loss": 4.931,
      "mean_token_accuracy": 0.20832182317972184,
      "num_tokens": 65224612.0,
      "step": 37600
    },
    {
      "entropy": 5.825767755508423,
      "epoch": 2.92581209881346,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041454847634054805,
      "loss": 4.9967,
      "mean_token_accuracy": 0.2157480537891388,
      "num_tokens": 65233389.0,
      "step": 37605
    },
    {
      "entropy": 5.841389226913452,
      "epoch": 2.926201128185178,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041452673430160923,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2106820434331894,
      "num_tokens": 65242432.0,
      "step": 37610
    },
    {
      "entropy": 5.706302642822266,
      "epoch": 2.9265901575568956,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004145049901455744,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2087727576494217,
      "num_tokens": 65250161.0,
      "step": 37615
    },
    {
      "entropy": 5.7493185043334964,
      "epoch": 2.9269791869286133,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041448324387277337,
      "loss": 5.0193,
      "mean_token_accuracy": 0.20658731162548066,
      "num_tokens": 65258660.0,
      "step": 37620
    },
    {
      "entropy": 5.827035284042358,
      "epoch": 2.927368216300331,
      "grad_norm": 1.25,
      "learning_rate": 0.00041446149548353606,
      "loss": 4.9994,
      "mean_token_accuracy": 0.2047867864370346,
      "num_tokens": 65267297.0,
      "step": 37625
    },
    {
      "entropy": 5.852462339401245,
      "epoch": 2.927757245672048,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004144397449781926,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20393924713134765,
      "num_tokens": 65276064.0,
      "step": 37630
    },
    {
      "entropy": 5.8386465549469,
      "epoch": 2.928146275043766,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004144179923570729,
      "loss": 5.019,
      "mean_token_accuracy": 0.2092908352613449,
      "num_tokens": 65284624.0,
      "step": 37635
    },
    {
      "entropy": 5.754473066329956,
      "epoch": 2.928535304415483,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004143962376205071,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21091643422842027,
      "num_tokens": 65293318.0,
      "step": 37640
    },
    {
      "entropy": 5.8159003257751465,
      "epoch": 2.928924333787201,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041437448076882526,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20696722865104675,
      "num_tokens": 65301379.0,
      "step": 37645
    },
    {
      "entropy": 5.821275615692139,
      "epoch": 2.9293133631589185,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041435272180235757,
      "loss": 5.0799,
      "mean_token_accuracy": 0.20018763840198517,
      "num_tokens": 65309722.0,
      "step": 37650
    },
    {
      "entropy": 5.751334953308105,
      "epoch": 2.9297023925306362,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000414330960721434,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21123647093772888,
      "num_tokens": 65318616.0,
      "step": 37655
    },
    {
      "entropy": 5.781906986236573,
      "epoch": 2.9300914219023535,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041430919752638497,
      "loss": 4.9616,
      "mean_token_accuracy": 0.21338335424661636,
      "num_tokens": 65327046.0,
      "step": 37660
    },
    {
      "entropy": 5.760299015045166,
      "epoch": 2.930480451274071,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004142874322175406,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20578066408634185,
      "num_tokens": 65335032.0,
      "step": 37665
    },
    {
      "entropy": 5.796276521682739,
      "epoch": 2.930869480645789,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041426566479523116,
      "loss": 5.0287,
      "mean_token_accuracy": 0.1976190209388733,
      "num_tokens": 65343573.0,
      "step": 37670
    },
    {
      "entropy": 5.734641885757446,
      "epoch": 2.931258510017506,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041424389525978683,
      "loss": 4.8747,
      "mean_token_accuracy": 0.21588744670152665,
      "num_tokens": 65352660.0,
      "step": 37675
    },
    {
      "entropy": 5.808008193969727,
      "epoch": 2.931647539389224,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004142221236115381,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20414777398109435,
      "num_tokens": 65362938.0,
      "step": 37680
    },
    {
      "entropy": 5.772867965698242,
      "epoch": 2.9320365687609415,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004142003498508152,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21310071498155594,
      "num_tokens": 65372000.0,
      "step": 37685
    },
    {
      "entropy": 5.739694261550904,
      "epoch": 2.9324255981326592,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004141785739779486,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2118447855114937,
      "num_tokens": 65379389.0,
      "step": 37690
    },
    {
      "entropy": 5.71614179611206,
      "epoch": 2.9328146275043765,
      "grad_norm": 1.25,
      "learning_rate": 0.00041415679599326853,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20234861075878144,
      "num_tokens": 65388335.0,
      "step": 37695
    },
    {
      "entropy": 5.790607643127442,
      "epoch": 2.933203656876094,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041413501589710567,
      "loss": 5.0679,
      "mean_token_accuracy": 0.2057709202170372,
      "num_tokens": 65397280.0,
      "step": 37700
    },
    {
      "entropy": 5.856473970413208,
      "epoch": 2.9335926862478114,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004141132336897904,
      "loss": 5.1315,
      "mean_token_accuracy": 0.19908856898546218,
      "num_tokens": 65406600.0,
      "step": 37705
    },
    {
      "entropy": 5.779257011413574,
      "epoch": 2.933981715619529,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004140914493716533,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20932478904724122,
      "num_tokens": 65415642.0,
      "step": 37710
    },
    {
      "entropy": 5.787950611114502,
      "epoch": 2.934370744991247,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004140696629430248,
      "loss": 4.9855,
      "mean_token_accuracy": 0.20670025497674943,
      "num_tokens": 65423977.0,
      "step": 37715
    },
    {
      "entropy": 5.721336221694946,
      "epoch": 2.9347597743629645,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041404787440423534,
      "loss": 4.991,
      "mean_token_accuracy": 0.20898325741291046,
      "num_tokens": 65432567.0,
      "step": 37720
    },
    {
      "entropy": 5.797557687759399,
      "epoch": 2.9351488037346822,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041402608375561595,
      "loss": 5.014,
      "mean_token_accuracy": 0.2073627233505249,
      "num_tokens": 65441341.0,
      "step": 37725
    },
    {
      "entropy": 5.721837568283081,
      "epoch": 2.9355378331063995,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041400429099749687,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21930188834667205,
      "num_tokens": 65449612.0,
      "step": 37730
    },
    {
      "entropy": 5.757532262802124,
      "epoch": 2.935926862478117,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000413982496130209,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19885763227939607,
      "num_tokens": 65457216.0,
      "step": 37735
    },
    {
      "entropy": 5.709102058410645,
      "epoch": 2.9363158918498344,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041396069915408294,
      "loss": 4.9123,
      "mean_token_accuracy": 0.20788934975862502,
      "num_tokens": 65465420.0,
      "step": 37740
    },
    {
      "entropy": 5.850804090499878,
      "epoch": 2.936704921221552,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041393890006944945,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2008939266204834,
      "num_tokens": 65474608.0,
      "step": 37745
    },
    {
      "entropy": 5.818904685974121,
      "epoch": 2.93709395059327,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004139170988766393,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21055236011743544,
      "num_tokens": 65483204.0,
      "step": 37750
    },
    {
      "entropy": 5.731201982498169,
      "epoch": 2.9374829799649875,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004138952955759833,
      "loss": 4.9173,
      "mean_token_accuracy": 0.21861817091703414,
      "num_tokens": 65491472.0,
      "step": 37755
    },
    {
      "entropy": 5.824015283584595,
      "epoch": 2.937872009336705,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041387349016781225,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2036407172679901,
      "num_tokens": 65500018.0,
      "step": 37760
    },
    {
      "entropy": 5.788583087921142,
      "epoch": 2.9382610387084225,
      "grad_norm": 1.234375,
      "learning_rate": 0.000413851682652457,
      "loss": 5.072,
      "mean_token_accuracy": 0.2095557376742363,
      "num_tokens": 65508845.0,
      "step": 37765
    },
    {
      "entropy": 5.79561128616333,
      "epoch": 2.93865006808014,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004138298730302485,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21410434395074845,
      "num_tokens": 65517696.0,
      "step": 37770
    },
    {
      "entropy": 5.745363759994507,
      "epoch": 2.9390390974518574,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004138080613015176,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20825862735509873,
      "num_tokens": 65526563.0,
      "step": 37775
    },
    {
      "entropy": 5.810450029373169,
      "epoch": 2.939428126823575,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041378624746659536,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2011308565735817,
      "num_tokens": 65535719.0,
      "step": 37780
    },
    {
      "entropy": 5.851435565948487,
      "epoch": 2.939817156195293,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041376443152581265,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19634182751178741,
      "num_tokens": 65543823.0,
      "step": 37785
    },
    {
      "entropy": 5.799710273742676,
      "epoch": 2.9402061855670105,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041374261347950064,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21321020871400834,
      "num_tokens": 65551923.0,
      "step": 37790
    },
    {
      "entropy": 5.793044757843018,
      "epoch": 2.9405952149387278,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004137207933279903,
      "loss": 4.9733,
      "mean_token_accuracy": 0.2130755975842476,
      "num_tokens": 65560552.0,
      "step": 37795
    },
    {
      "entropy": 5.796958589553833,
      "epoch": 2.9409842443104455,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041369897107161266,
      "loss": 4.994,
      "mean_token_accuracy": 0.20644956678152085,
      "num_tokens": 65568897.0,
      "step": 37800
    },
    {
      "entropy": 5.774103736877441,
      "epoch": 2.941373273682163,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041367714671069897,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20993162393569947,
      "num_tokens": 65577611.0,
      "step": 37805
    },
    {
      "entropy": 5.790308952331543,
      "epoch": 2.9417623030538804,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041365532024558024,
      "loss": 4.9779,
      "mean_token_accuracy": 0.20606066435575485,
      "num_tokens": 65586284.0,
      "step": 37810
    },
    {
      "entropy": 5.82007212638855,
      "epoch": 2.942151332425598,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004136334916765877,
      "loss": 5.0141,
      "mean_token_accuracy": 0.20786223113536834,
      "num_tokens": 65595065.0,
      "step": 37815
    },
    {
      "entropy": 5.846232461929321,
      "epoch": 2.942540361797316,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041361166100405264,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20374172925949097,
      "num_tokens": 65604478.0,
      "step": 37820
    },
    {
      "entropy": 5.760382795333863,
      "epoch": 2.9429293911690335,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004135898282283062,
      "loss": 5.035,
      "mean_token_accuracy": 0.20741840749979018,
      "num_tokens": 65613473.0,
      "step": 37825
    },
    {
      "entropy": 5.7852343082427975,
      "epoch": 2.9433184205407508,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004135679933496797,
      "loss": 4.962,
      "mean_token_accuracy": 0.211484332382679,
      "num_tokens": 65622498.0,
      "step": 37830
    },
    {
      "entropy": 5.770973110198975,
      "epoch": 2.9437074499124685,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004135461563685045,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20100015997886658,
      "num_tokens": 65630939.0,
      "step": 37835
    },
    {
      "entropy": 5.82726731300354,
      "epoch": 2.9440964792841857,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041352431728511194,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20613081455230714,
      "num_tokens": 65639667.0,
      "step": 37840
    },
    {
      "entropy": 5.861242341995239,
      "epoch": 2.9444855086559034,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004135024760998333,
      "loss": 5.0004,
      "mean_token_accuracy": 0.21026089340448378,
      "num_tokens": 65648413.0,
      "step": 37845
    },
    {
      "entropy": 5.864446401596069,
      "epoch": 2.944874538027621,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041348063281299995,
      "loss": 5.079,
      "mean_token_accuracy": 0.20238011479377746,
      "num_tokens": 65657486.0,
      "step": 37850
    },
    {
      "entropy": 5.767792654037476,
      "epoch": 2.945263567399339,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041345878742494346,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2084062486886978,
      "num_tokens": 65666328.0,
      "step": 37855
    },
    {
      "entropy": 5.723753213882446,
      "epoch": 2.945652596771056,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041343693993599535,
      "loss": 4.9889,
      "mean_token_accuracy": 0.201590134203434,
      "num_tokens": 65674107.0,
      "step": 37860
    },
    {
      "entropy": 5.711829805374146,
      "epoch": 2.9460416261427738,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004134150903464869,
      "loss": 4.9179,
      "mean_token_accuracy": 0.205091655254364,
      "num_tokens": 65682534.0,
      "step": 37865
    },
    {
      "entropy": 5.85307559967041,
      "epoch": 2.9464306555144915,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004133932386567498,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1932014286518097,
      "num_tokens": 65690781.0,
      "step": 37870
    },
    {
      "entropy": 5.829770040512085,
      "epoch": 2.9468196848862087,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004133713848671156,
      "loss": 4.9486,
      "mean_token_accuracy": 0.2112462416291237,
      "num_tokens": 65698668.0,
      "step": 37875
    },
    {
      "entropy": 5.805991697311401,
      "epoch": 2.9472087142579264,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004133495289779159,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2082289695739746,
      "num_tokens": 65706568.0,
      "step": 37880
    },
    {
      "entropy": 5.746881914138794,
      "epoch": 2.947597743629644,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004133276709894823,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20366055965423585,
      "num_tokens": 65714652.0,
      "step": 37885
    },
    {
      "entropy": 5.7513134479522705,
      "epoch": 2.947986773001362,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004133058109021464,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21560180932283401,
      "num_tokens": 65723296.0,
      "step": 37890
    },
    {
      "entropy": 5.822782754898071,
      "epoch": 2.948375802373079,
      "grad_norm": 1.25,
      "learning_rate": 0.00041328394871624003,
      "loss": 5.0679,
      "mean_token_accuracy": 0.19745966643095017,
      "num_tokens": 65732257.0,
      "step": 37895
    },
    {
      "entropy": 5.839972972869873,
      "epoch": 2.9487648317447968,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004132620844320948,
      "loss": 4.9612,
      "mean_token_accuracy": 0.20625888109207152,
      "num_tokens": 65740152.0,
      "step": 37900
    },
    {
      "entropy": 5.779817676544189,
      "epoch": 2.9491538611165145,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004132402180500425,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20470497012138367,
      "num_tokens": 65748927.0,
      "step": 37905
    },
    {
      "entropy": 5.7853928565979,
      "epoch": 2.9495428904882317,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004132183495704149,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20228881984949112,
      "num_tokens": 65756949.0,
      "step": 37910
    },
    {
      "entropy": 5.888367605209351,
      "epoch": 2.9499319198599494,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004131964789935439,
      "loss": 5.1302,
      "mean_token_accuracy": 0.19500030130147933,
      "num_tokens": 65765076.0,
      "step": 37915
    },
    {
      "entropy": 5.77508955001831,
      "epoch": 2.950320949231667,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004131746063197612,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20988013744354247,
      "num_tokens": 65773581.0,
      "step": 37920
    },
    {
      "entropy": 5.855573892593384,
      "epoch": 2.950709978603385,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004131527315493989,
      "loss": 5.1179,
      "mean_token_accuracy": 0.20234867036342621,
      "num_tokens": 65782312.0,
      "step": 37925
    },
    {
      "entropy": 5.760341167449951,
      "epoch": 2.951099007975102,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041313085468278873,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21106754392385482,
      "num_tokens": 65790601.0,
      "step": 37930
    },
    {
      "entropy": 5.790761041641235,
      "epoch": 2.9514880373468197,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004131089757202627,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20278241038322448,
      "num_tokens": 65799393.0,
      "step": 37935
    },
    {
      "entropy": 5.818588209152222,
      "epoch": 2.951877066718537,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041308709466215276,
      "loss": 4.9917,
      "mean_token_accuracy": 0.20847802013158798,
      "num_tokens": 65808871.0,
      "step": 37940
    },
    {
      "entropy": 5.700021505355835,
      "epoch": 2.9522660960902547,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000413065211508791,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21825315952301025,
      "num_tokens": 65817075.0,
      "step": 37945
    },
    {
      "entropy": 5.741752815246582,
      "epoch": 2.9526551254619724,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041304332626050937,
      "loss": 4.9146,
      "mean_token_accuracy": 0.21415839791297914,
      "num_tokens": 65825081.0,
      "step": 37950
    },
    {
      "entropy": 5.777788829803467,
      "epoch": 2.95304415483369,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041302143891764,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21135697662830352,
      "num_tokens": 65833183.0,
      "step": 37955
    },
    {
      "entropy": 5.807670164108276,
      "epoch": 2.9534331842054073,
      "grad_norm": 1.234375,
      "learning_rate": 0.000412999549480515,
      "loss": 4.9209,
      "mean_token_accuracy": 0.20541346669197083,
      "num_tokens": 65841514.0,
      "step": 37960
    },
    {
      "entropy": 5.857275104522705,
      "epoch": 2.953822213577125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004129776579494665,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21420253366231917,
      "num_tokens": 65850433.0,
      "step": 37965
    },
    {
      "entropy": 5.834450626373291,
      "epoch": 2.9542112429488427,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004129557643248265,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2018587052822113,
      "num_tokens": 65859807.0,
      "step": 37970
    },
    {
      "entropy": 5.798654413223266,
      "epoch": 2.95460027232056,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004129338686069276,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21580596268177032,
      "num_tokens": 65867901.0,
      "step": 37975
    },
    {
      "entropy": 5.776148939132691,
      "epoch": 2.9549893016922777,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004129119707961016,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20699337422847747,
      "num_tokens": 65876660.0,
      "step": 37980
    },
    {
      "entropy": 5.8715667724609375,
      "epoch": 2.9553783310639954,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000412890070892681,
      "loss": 5.0476,
      "mean_token_accuracy": 0.20491838753223418,
      "num_tokens": 65885258.0,
      "step": 37985
    },
    {
      "entropy": 5.827898788452148,
      "epoch": 2.955767360435713,
      "grad_norm": 1.3125,
      "learning_rate": 0.000412868168896998,
      "loss": 5.086,
      "mean_token_accuracy": 0.19763365983963013,
      "num_tokens": 65894260.0,
      "step": 37990
    },
    {
      "entropy": 5.837190628051758,
      "epoch": 2.9561563898074303,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004128462648093851,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20563557147979736,
      "num_tokens": 65903000.0,
      "step": 37995
    },
    {
      "entropy": 5.806035661697388,
      "epoch": 2.956545419179148,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004128243586301745,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20684773325920106,
      "num_tokens": 65911550.0,
      "step": 38000
    },
    {
      "entropy": 5.843602705001831,
      "epoch": 2.9569344485508657,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004128024503596986,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21079884469509125,
      "num_tokens": 65920091.0,
      "step": 38005
    },
    {
      "entropy": 5.773313093185425,
      "epoch": 2.957323477922583,
      "grad_norm": 1.25,
      "learning_rate": 0.00041278053999828986,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2080454409122467,
      "num_tokens": 65929313.0,
      "step": 38010
    },
    {
      "entropy": 5.769234991073608,
      "epoch": 2.9577125072943007,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004127586275462806,
      "loss": 5.0554,
      "mean_token_accuracy": 0.2013596475124359,
      "num_tokens": 65938407.0,
      "step": 38015
    },
    {
      "entropy": 5.826012706756591,
      "epoch": 2.9581015366660184,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041273671300400363,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20681959688663482,
      "num_tokens": 65946644.0,
      "step": 38020
    },
    {
      "entropy": 5.822836875915527,
      "epoch": 2.958490566037736,
      "grad_norm": 1.25,
      "learning_rate": 0.00041271479637179116,
      "loss": 5.0728,
      "mean_token_accuracy": 0.20069655030965805,
      "num_tokens": 65955695.0,
      "step": 38025
    },
    {
      "entropy": 5.8413115501403805,
      "epoch": 2.9588795954094533,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004126928776499759,
      "loss": 5.107,
      "mean_token_accuracy": 0.19351697862148284,
      "num_tokens": 65964326.0,
      "step": 38030
    },
    {
      "entropy": 5.746098566055298,
      "epoch": 2.959268624781171,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004126709568388903,
      "loss": 4.9195,
      "mean_token_accuracy": 0.20924546271562577,
      "num_tokens": 65972986.0,
      "step": 38035
    },
    {
      "entropy": 5.764552164077759,
      "epoch": 2.9596576541528883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004126490339388672,
      "loss": 4.9754,
      "mean_token_accuracy": 0.20648670792579651,
      "num_tokens": 65980829.0,
      "step": 38040
    },
    {
      "entropy": 5.84832911491394,
      "epoch": 2.960046683524606,
      "grad_norm": 1.359375,
      "learning_rate": 0.000412627108950239,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21136318594217302,
      "num_tokens": 65989412.0,
      "step": 38045
    },
    {
      "entropy": 5.745075035095215,
      "epoch": 2.9604357128963237,
      "grad_norm": 1.25,
      "learning_rate": 0.0004126051818733385,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21510806381702424,
      "num_tokens": 65998887.0,
      "step": 38050
    },
    {
      "entropy": 5.785877084732055,
      "epoch": 2.9608247422680414,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004125832527084984,
      "loss": 5.0343,
      "mean_token_accuracy": 0.2103905126452446,
      "num_tokens": 66007266.0,
      "step": 38055
    },
    {
      "entropy": 5.770119953155517,
      "epoch": 2.961213771639759,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041256132145605143,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2103514015674591,
      "num_tokens": 66015968.0,
      "step": 38060
    },
    {
      "entropy": 5.814166021347046,
      "epoch": 2.9616028010114763,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004125393881163304,
      "loss": 5.0554,
      "mean_token_accuracy": 0.19837448745965958,
      "num_tokens": 66025128.0,
      "step": 38065
    },
    {
      "entropy": 5.783906698226929,
      "epoch": 2.961991830383194,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041251745268966797,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21159233748912812,
      "num_tokens": 66033730.0,
      "step": 38070
    },
    {
      "entropy": 5.702244520187378,
      "epoch": 2.9623808597549113,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004124955151763971,
      "loss": 4.895,
      "mean_token_accuracy": 0.21426949203014373,
      "num_tokens": 66042711.0,
      "step": 38075
    },
    {
      "entropy": 5.824656343460083,
      "epoch": 2.962769889126629,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004124735755768506,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2093021899461746,
      "num_tokens": 66051107.0,
      "step": 38080
    },
    {
      "entropy": 5.771162271499634,
      "epoch": 2.9631589184983467,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004124516338913615,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21530238687992095,
      "num_tokens": 66059532.0,
      "step": 38085
    },
    {
      "entropy": 5.774539041519165,
      "epoch": 2.9635479478700644,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004124296901202626,
      "loss": 5.007,
      "mean_token_accuracy": 0.1999477580189705,
      "num_tokens": 66068034.0,
      "step": 38090
    },
    {
      "entropy": 5.787468910217285,
      "epoch": 2.9639369772417816,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041240774426388685,
      "loss": 5.0556,
      "mean_token_accuracy": 0.2033907249569893,
      "num_tokens": 66077403.0,
      "step": 38095
    },
    {
      "entropy": 5.794546365737915,
      "epoch": 2.9643260066134993,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004123857963225673,
      "loss": 5.0254,
      "mean_token_accuracy": 0.19726659506559371,
      "num_tokens": 66085930.0,
      "step": 38100
    },
    {
      "entropy": 5.832353258132935,
      "epoch": 2.964715035985217,
      "grad_norm": 1.359375,
      "learning_rate": 0.000412363846296637,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20918936878442765,
      "num_tokens": 66094671.0,
      "step": 38105
    },
    {
      "entropy": 5.816906213760376,
      "epoch": 2.9651040653569343,
      "grad_norm": 1.28125,
      "learning_rate": 0.000412341894186429,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20822850465774537,
      "num_tokens": 66103250.0,
      "step": 38110
    },
    {
      "entropy": 5.761331415176391,
      "epoch": 2.965493094728652,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004123199399922763,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20731908828020096,
      "num_tokens": 66111073.0,
      "step": 38115
    },
    {
      "entropy": 5.739909029006958,
      "epoch": 2.9658821241003697,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041229798371451203,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2095324292778969,
      "num_tokens": 66119514.0,
      "step": 38120
    },
    {
      "entropy": 5.76384220123291,
      "epoch": 2.9662711534720874,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004122760253534695,
      "loss": 4.9601,
      "mean_token_accuracy": 0.20746271908283234,
      "num_tokens": 66128804.0,
      "step": 38125
    },
    {
      "entropy": 5.803957557678222,
      "epoch": 2.9666601828438046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004122540649094817,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21370601803064346,
      "num_tokens": 66137683.0,
      "step": 38130
    },
    {
      "entropy": 5.832731056213379,
      "epoch": 2.9670492122155223,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041223210238288196,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20532537400722503,
      "num_tokens": 66147135.0,
      "step": 38135
    },
    {
      "entropy": 5.819063568115235,
      "epoch": 2.9674382415872396,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004122101377740035,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21060555726289748,
      "num_tokens": 66155534.0,
      "step": 38140
    },
    {
      "entropy": 5.7921661853790285,
      "epoch": 2.9678272709589573,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004121881710831796,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2114035815000534,
      "num_tokens": 66164413.0,
      "step": 38145
    },
    {
      "entropy": 5.788314533233643,
      "epoch": 2.968216300330675,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004121662023107435,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20435655266046523,
      "num_tokens": 66173469.0,
      "step": 38150
    },
    {
      "entropy": 5.745444488525391,
      "epoch": 2.9686053297023927,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004121442314570286,
      "loss": 4.9403,
      "mean_token_accuracy": 0.20723003894090652,
      "num_tokens": 66182183.0,
      "step": 38155
    },
    {
      "entropy": 5.800443315505982,
      "epoch": 2.9689943590741104,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041212225852236834,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21003050357103348,
      "num_tokens": 66190647.0,
      "step": 38160
    },
    {
      "entropy": 5.691683197021485,
      "epoch": 2.9693833884458276,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004121002835070961,
      "loss": 4.8687,
      "mean_token_accuracy": 0.21443312615156174,
      "num_tokens": 66198536.0,
      "step": 38165
    },
    {
      "entropy": 5.802857065200806,
      "epoch": 2.9697724178175453,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004120783064115452,
      "loss": 5.1497,
      "mean_token_accuracy": 0.200189246237278,
      "num_tokens": 66207772.0,
      "step": 38170
    },
    {
      "entropy": 5.816162109375,
      "epoch": 2.9701614471892626,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004120563272360492,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2008711263537407,
      "num_tokens": 66216862.0,
      "step": 38175
    },
    {
      "entropy": 5.840185546875,
      "epoch": 2.9705504765609803,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004120343459809416,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20995634347200393,
      "num_tokens": 66225817.0,
      "step": 38180
    },
    {
      "entropy": 5.8212841033935545,
      "epoch": 2.970939505932698,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004120123626465559,
      "loss": 5.043,
      "mean_token_accuracy": 0.20360663831233977,
      "num_tokens": 66234894.0,
      "step": 38185
    },
    {
      "entropy": 5.762087059020996,
      "epoch": 2.9713285353044157,
      "grad_norm": 1.25,
      "learning_rate": 0.00041199037723322566,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21042720526456832,
      "num_tokens": 66243800.0,
      "step": 38190
    },
    {
      "entropy": 5.772845792770386,
      "epoch": 2.971717564676133,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041196838974128446,
      "loss": 4.9205,
      "mean_token_accuracy": 0.20566324144601822,
      "num_tokens": 66253192.0,
      "step": 38195
    },
    {
      "entropy": 5.74368143081665,
      "epoch": 2.9721065940478506,
      "grad_norm": 1.21875,
      "learning_rate": 0.000411946400171066,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20748110264539718,
      "num_tokens": 66261775.0,
      "step": 38200
    },
    {
      "entropy": 5.731686210632324,
      "epoch": 2.9724956234195683,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041192440852290385,
      "loss": 4.988,
      "mean_token_accuracy": 0.20611457526683807,
      "num_tokens": 66271069.0,
      "step": 38205
    },
    {
      "entropy": 5.864109659194947,
      "epoch": 2.9728846527912856,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004119024147971318,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20084852427244188,
      "num_tokens": 66280021.0,
      "step": 38210
    },
    {
      "entropy": 5.765340757369995,
      "epoch": 2.9732736821630033,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041188041899408345,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21009047329425812,
      "num_tokens": 66288782.0,
      "step": 38215
    },
    {
      "entropy": 5.81737322807312,
      "epoch": 2.973662711534721,
      "grad_norm": 1.375,
      "learning_rate": 0.0004118584211140926,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2023215502500534,
      "num_tokens": 66297716.0,
      "step": 38220
    },
    {
      "entropy": 5.901015949249268,
      "epoch": 2.9740517409064386,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041183642115749316,
      "loss": 5.0257,
      "mean_token_accuracy": 0.20483824610710144,
      "num_tokens": 66305977.0,
      "step": 38225
    },
    {
      "entropy": 5.838449144363404,
      "epoch": 2.974440770278156,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041181441912461873,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20575713664293288,
      "num_tokens": 66314344.0,
      "step": 38230
    },
    {
      "entropy": 5.740027189254761,
      "epoch": 2.9748297996498736,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004117924150158032,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2074679970741272,
      "num_tokens": 66322776.0,
      "step": 38235
    },
    {
      "entropy": 5.814639949798584,
      "epoch": 2.9752188290215913,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041177040883138064,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21037317663431168,
      "num_tokens": 66331161.0,
      "step": 38240
    },
    {
      "entropy": 5.802111864089966,
      "epoch": 2.9756078583933085,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041174840057168474,
      "loss": 5.0286,
      "mean_token_accuracy": 0.20305119454860687,
      "num_tokens": 66339579.0,
      "step": 38245
    },
    {
      "entropy": 5.669948959350586,
      "epoch": 2.9759968877650262,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004117263902370495,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2093421846628189,
      "num_tokens": 66348716.0,
      "step": 38250
    },
    {
      "entropy": 5.807838821411133,
      "epoch": 2.976385917136744,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041170437782780896,
      "loss": 4.9825,
      "mean_token_accuracy": 0.20322963893413543,
      "num_tokens": 66357546.0,
      "step": 38255
    },
    {
      "entropy": 5.757318592071533,
      "epoch": 2.9767749465084616,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000411682363344297,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22234776765108108,
      "num_tokens": 66366024.0,
      "step": 38260
    },
    {
      "entropy": 5.766235733032227,
      "epoch": 2.977163975880179,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041166034678684777,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21150738298892974,
      "num_tokens": 66374413.0,
      "step": 38265
    },
    {
      "entropy": 5.772994899749756,
      "epoch": 2.9775530052518966,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004116383281557953,
      "loss": 5.0273,
      "mean_token_accuracy": 0.19758633077144622,
      "num_tokens": 66383131.0,
      "step": 38270
    },
    {
      "entropy": 5.80317759513855,
      "epoch": 2.977942034623614,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041161630745147374,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19683407843112946,
      "num_tokens": 66392227.0,
      "step": 38275
    },
    {
      "entropy": 5.790156555175781,
      "epoch": 2.9783310639953315,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004115942846742171,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2071742370724678,
      "num_tokens": 66401384.0,
      "step": 38280
    },
    {
      "entropy": 5.767789649963379,
      "epoch": 2.9787200933670492,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004115722598243596,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21277025192975998,
      "num_tokens": 66409936.0,
      "step": 38285
    },
    {
      "entropy": 5.751672315597534,
      "epoch": 2.979109122738767,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041155023290223537,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21406008899211884,
      "num_tokens": 66418444.0,
      "step": 38290
    },
    {
      "entropy": 5.754353857040405,
      "epoch": 2.979498152110484,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004115282039081788,
      "loss": 4.999,
      "mean_token_accuracy": 0.20769090354442596,
      "num_tokens": 66427211.0,
      "step": 38295
    },
    {
      "entropy": 5.757352304458618,
      "epoch": 2.979887181482202,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041150617284252405,
      "loss": 4.967,
      "mean_token_accuracy": 0.21303538531064986,
      "num_tokens": 66435767.0,
      "step": 38300
    },
    {
      "entropy": 5.869857549667358,
      "epoch": 2.9802762108539196,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004114841397056053,
      "loss": 5.0158,
      "mean_token_accuracy": 0.20011367350816728,
      "num_tokens": 66443647.0,
      "step": 38305
    },
    {
      "entropy": 5.775699710845947,
      "epoch": 2.980665240225637,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041146210449775704,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21393524259328842,
      "num_tokens": 66452011.0,
      "step": 38310
    },
    {
      "entropy": 5.836452627182007,
      "epoch": 2.9810542695973545,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004114400672193136,
      "loss": 5.058,
      "mean_token_accuracy": 0.2017641082406044,
      "num_tokens": 66460151.0,
      "step": 38315
    },
    {
      "entropy": 5.833781099319458,
      "epoch": 2.9814432989690722,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041141802787060923,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21065867990255355,
      "num_tokens": 66468409.0,
      "step": 38320
    },
    {
      "entropy": 5.789581775665283,
      "epoch": 2.98183232834079,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004113959864519784,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20147260874509812,
      "num_tokens": 66477765.0,
      "step": 38325
    },
    {
      "entropy": 5.822074031829834,
      "epoch": 2.982221357712507,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004113739429637556,
      "loss": 5.0938,
      "mean_token_accuracy": 0.19924278557300568,
      "num_tokens": 66486726.0,
      "step": 38330
    },
    {
      "entropy": 5.862177467346191,
      "epoch": 2.982610387084225,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004113518974062753,
      "loss": 5.0686,
      "mean_token_accuracy": 0.19698427021503448,
      "num_tokens": 66495994.0,
      "step": 38335
    },
    {
      "entropy": 5.885537958145141,
      "epoch": 2.9829994164559426,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000411329849779872,
      "loss": 5.0442,
      "mean_token_accuracy": 0.200996558368206,
      "num_tokens": 66504515.0,
      "step": 38340
    },
    {
      "entropy": 5.771761703491211,
      "epoch": 2.98338844582766,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041130780008488015,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21198905855417252,
      "num_tokens": 66513637.0,
      "step": 38345
    },
    {
      "entropy": 5.75206036567688,
      "epoch": 2.9837774751993775,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004112857483216344,
      "loss": 4.9222,
      "mean_token_accuracy": 0.21268266886472703,
      "num_tokens": 66521453.0,
      "step": 38350
    },
    {
      "entropy": 5.6420519828796385,
      "epoch": 2.9841665045710952,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041126369449046935,
      "loss": 4.8017,
      "mean_token_accuracy": 0.22639191746711732,
      "num_tokens": 66530888.0,
      "step": 38355
    },
    {
      "entropy": 5.7975578784942625,
      "epoch": 2.984555533942813,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041124163859171967,
      "loss": 5.0683,
      "mean_token_accuracy": 0.2045467644929886,
      "num_tokens": 66538758.0,
      "step": 38360
    },
    {
      "entropy": 5.837550973892212,
      "epoch": 2.98494456331453,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004112195806257199,
      "loss": 5.0611,
      "mean_token_accuracy": 0.19752038717269899,
      "num_tokens": 66546981.0,
      "step": 38365
    },
    {
      "entropy": 5.787695598602295,
      "epoch": 2.985333592686248,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004111975205928049,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20252190083265303,
      "num_tokens": 66555375.0,
      "step": 38370
    },
    {
      "entropy": 5.800839233398437,
      "epoch": 2.985722622057965,
      "grad_norm": 1.25,
      "learning_rate": 0.00041117545849330924,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2106126666069031,
      "num_tokens": 66564581.0,
      "step": 38375
    },
    {
      "entropy": 5.777060699462891,
      "epoch": 2.986111651429683,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041115339432756776,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20744550973176956,
      "num_tokens": 66573876.0,
      "step": 38380
    },
    {
      "entropy": 5.8014768123626705,
      "epoch": 2.9865006808014005,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004111313280959152,
      "loss": 5.0315,
      "mean_token_accuracy": 0.2167380303144455,
      "num_tokens": 66582039.0,
      "step": 38385
    },
    {
      "entropy": 5.77006254196167,
      "epoch": 2.986889710173118,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041110925979868647,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22029772102832795,
      "num_tokens": 66590188.0,
      "step": 38390
    },
    {
      "entropy": 5.763982200622559,
      "epoch": 2.9872787395448355,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004110871894362163,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20857761055231094,
      "num_tokens": 66598889.0,
      "step": 38395
    },
    {
      "entropy": 5.774786186218262,
      "epoch": 2.987667768916553,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041106511700883975,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20654350519180298,
      "num_tokens": 66606909.0,
      "step": 38400
    },
    {
      "entropy": 5.764012241363526,
      "epoch": 2.988056798288271,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041104304251689147,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20620758086442947,
      "num_tokens": 66615644.0,
      "step": 38405
    },
    {
      "entropy": 5.882189512252808,
      "epoch": 2.988445827659988,
      "grad_norm": 1.25,
      "learning_rate": 0.00041102096596070666,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20396297723054885,
      "num_tokens": 66624466.0,
      "step": 38410
    },
    {
      "entropy": 5.908952426910401,
      "epoch": 2.988834857031706,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041099888734062013,
      "loss": 5.145,
      "mean_token_accuracy": 0.1980953112244606,
      "num_tokens": 66633907.0,
      "step": 38415
    },
    {
      "entropy": 5.792241764068604,
      "epoch": 2.9892238864034235,
      "grad_norm": 1.296875,
      "learning_rate": 0.000410976806656967,
      "loss": 5.083,
      "mean_token_accuracy": 0.20644789189100266,
      "num_tokens": 66643652.0,
      "step": 38420
    },
    {
      "entropy": 5.821304988861084,
      "epoch": 2.989612915775141,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004109547239100822,
      "loss": 5.0426,
      "mean_token_accuracy": 0.19958211481571198,
      "num_tokens": 66652440.0,
      "step": 38425
    },
    {
      "entropy": 5.829813766479492,
      "epoch": 2.9900019451468585,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004109326391003009,
      "loss": 4.9826,
      "mean_token_accuracy": 0.20672493427991867,
      "num_tokens": 66661660.0,
      "step": 38430
    },
    {
      "entropy": 5.861014747619629,
      "epoch": 2.990390974518576,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041091055222795814,
      "loss": 4.9895,
      "mean_token_accuracy": 0.2067125454545021,
      "num_tokens": 66670289.0,
      "step": 38435
    },
    {
      "entropy": 5.790329647064209,
      "epoch": 2.990780003890294,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004108884632933891,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21383788138628007,
      "num_tokens": 66679777.0,
      "step": 38440
    },
    {
      "entropy": 5.8540873527526855,
      "epoch": 2.991169033262011,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041086637229692893,
      "loss": 5.0415,
      "mean_token_accuracy": 0.20178776383399963,
      "num_tokens": 66689264.0,
      "step": 38445
    },
    {
      "entropy": 5.828787708282471,
      "epoch": 2.991558062633729,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041084427923891275,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20440392345190048,
      "num_tokens": 66697632.0,
      "step": 38450
    },
    {
      "entropy": 5.795511054992676,
      "epoch": 2.9919470920054465,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041082218411967586,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2096043199300766,
      "num_tokens": 66706312.0,
      "step": 38455
    },
    {
      "entropy": 5.834311676025391,
      "epoch": 2.992336121377164,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041080008693955355,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20693301856517793,
      "num_tokens": 66714322.0,
      "step": 38460
    },
    {
      "entropy": 5.758828163146973,
      "epoch": 2.9927251507488815,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004107779876988811,
      "loss": 4.9818,
      "mean_token_accuracy": 0.20128415077924727,
      "num_tokens": 66723039.0,
      "step": 38465
    },
    {
      "entropy": 5.761122322082519,
      "epoch": 2.993114180120599,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041075588639799375,
      "loss": 4.9733,
      "mean_token_accuracy": 0.2108639270067215,
      "num_tokens": 66731414.0,
      "step": 38470
    },
    {
      "entropy": 5.7801813125610355,
      "epoch": 2.9935032094923164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004107337830372269,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21190512478351592,
      "num_tokens": 66740664.0,
      "step": 38475
    },
    {
      "entropy": 5.792689085006714,
      "epoch": 2.993892238864034,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000410711677616916,
      "loss": 4.9458,
      "mean_token_accuracy": 0.2051389992237091,
      "num_tokens": 66749073.0,
      "step": 38480
    },
    {
      "entropy": 5.764163637161255,
      "epoch": 2.994281268235752,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004106895701373963,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2060171127319336,
      "num_tokens": 66757734.0,
      "step": 38485
    },
    {
      "entropy": 5.755512857437134,
      "epoch": 2.9946702976074695,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041066746059900336,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20302744656801225,
      "num_tokens": 66766699.0,
      "step": 38490
    },
    {
      "entropy": 5.799043321609497,
      "epoch": 2.995059326979187,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004106453490020727,
      "loss": 5.1025,
      "mean_token_accuracy": 0.19639278799295426,
      "num_tokens": 66775398.0,
      "step": 38495
    },
    {
      "entropy": 5.7222813129425045,
      "epoch": 2.9954483563509045,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00041062323534693973,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21909183263778687,
      "num_tokens": 66783157.0,
      "step": 38500
    },
    {
      "entropy": 5.715114498138428,
      "epoch": 2.995837385722622,
      "grad_norm": 1.375,
      "learning_rate": 0.00041060111963394004,
      "loss": 5.003,
      "mean_token_accuracy": 0.2069690302014351,
      "num_tokens": 66791293.0,
      "step": 38505
    },
    {
      "entropy": 5.6996386528015135,
      "epoch": 2.9962264150943394,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041057900186340917,
      "loss": 4.9012,
      "mean_token_accuracy": 0.21450725048780442,
      "num_tokens": 66800212.0,
      "step": 38510
    },
    {
      "entropy": 5.86918077468872,
      "epoch": 2.996615444466057,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041055688203568274,
      "loss": 5.0753,
      "mean_token_accuracy": 0.20039113014936447,
      "num_tokens": 66808050.0,
      "step": 38515
    },
    {
      "entropy": 5.8338268280029295,
      "epoch": 2.997004473837775,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004105347601510964,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21630944609642028,
      "num_tokens": 66816481.0,
      "step": 38520
    },
    {
      "entropy": 5.770847320556641,
      "epoch": 2.9973935032094925,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004105126362099859,
      "loss": 4.97,
      "mean_token_accuracy": 0.208756123483181,
      "num_tokens": 66825294.0,
      "step": 38525
    },
    {
      "entropy": 5.759933805465698,
      "epoch": 2.9977825325812097,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004104905102126867,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2119574189186096,
      "num_tokens": 66833818.0,
      "step": 38530
    },
    {
      "entropy": 5.824851083755493,
      "epoch": 2.9981715619529274,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004104683821595347,
      "loss": 5.056,
      "mean_token_accuracy": 0.20276401937007904,
      "num_tokens": 66842604.0,
      "step": 38535
    },
    {
      "entropy": 5.874629545211792,
      "epoch": 2.998560591324645,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041044625205086566,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20177406519651414,
      "num_tokens": 66850974.0,
      "step": 38540
    },
    {
      "entropy": 5.812937927246094,
      "epoch": 2.9989496206963624,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041042411988701535,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21352594047784806,
      "num_tokens": 66859671.0,
      "step": 38545
    },
    {
      "entropy": 5.812755680084228,
      "epoch": 2.99933865006808,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041040198566831944,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2062032088637352,
      "num_tokens": 66868864.0,
      "step": 38550
    },
    {
      "entropy": 5.747423553466797,
      "epoch": 2.999727679439798,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004103798493951141,
      "loss": 5.0028,
      "mean_token_accuracy": 0.20764863044023513,
      "num_tokens": 66877547.0,
      "step": 38555
    },
    {
      "entropy": 5.779676331414117,
      "epoch": 3.0000778058743434,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041035771106773483,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21295927464962006,
      "num_tokens": 66884854.0,
      "step": 38560
    },
    {
      "entropy": 5.82189621925354,
      "epoch": 3.000466835246061,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004103355706865179,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20353011637926102,
      "num_tokens": 66893963.0,
      "step": 38565
    },
    {
      "entropy": 5.744144058227539,
      "epoch": 3.0008558646177788,
      "grad_norm": 1.171875,
      "learning_rate": 0.000410313428251799,
      "loss": 4.931,
      "mean_token_accuracy": 0.2055003046989441,
      "num_tokens": 66902788.0,
      "step": 38570
    },
    {
      "entropy": 5.775751352310181,
      "epoch": 3.001244893989496,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041029128376391424,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20949519127607347,
      "num_tokens": 66912294.0,
      "step": 38575
    },
    {
      "entropy": 5.76895809173584,
      "epoch": 3.0016339233612137,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041026913722319954,
      "loss": 4.8804,
      "mean_token_accuracy": 0.21265367567539215,
      "num_tokens": 66920080.0,
      "step": 38580
    },
    {
      "entropy": 5.736233377456665,
      "epoch": 3.0020229527329314,
      "grad_norm": 1.28125,
      "learning_rate": 0.000410246988629991,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2075844094157219,
      "num_tokens": 66928404.0,
      "step": 38585
    },
    {
      "entropy": 5.826994752883911,
      "epoch": 3.002411982104649,
      "grad_norm": 1.25,
      "learning_rate": 0.00041022483798462466,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2071014776825905,
      "num_tokens": 66938040.0,
      "step": 38590
    },
    {
      "entropy": 5.824998807907105,
      "epoch": 3.0028010114763664,
      "grad_norm": 1.25,
      "learning_rate": 0.0004102026852874367,
      "loss": 4.9685,
      "mean_token_accuracy": 0.21312396973371506,
      "num_tokens": 66946605.0,
      "step": 38595
    },
    {
      "entropy": 5.7851190090179445,
      "epoch": 3.003190040848084,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041018053053876316,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20936891287565232,
      "num_tokens": 66955276.0,
      "step": 38600
    },
    {
      "entropy": 5.814677095413208,
      "epoch": 3.0035790702198017,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004101583737389402,
      "loss": 4.9549,
      "mean_token_accuracy": 0.21053484380245208,
      "num_tokens": 66963950.0,
      "step": 38605
    },
    {
      "entropy": 5.802759408950806,
      "epoch": 3.003968099591519,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041013621488830406,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21047796458005905,
      "num_tokens": 66972101.0,
      "step": 38610
    },
    {
      "entropy": 5.711282444000244,
      "epoch": 3.0043571289632367,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041011405398719096,
      "loss": 4.8744,
      "mean_token_accuracy": 0.21371332854032515,
      "num_tokens": 66981251.0,
      "step": 38615
    },
    {
      "entropy": 5.811426830291748,
      "epoch": 3.0047461583349544,
      "grad_norm": 1.25,
      "learning_rate": 0.00041009189103593716,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20761600285768508,
      "num_tokens": 66990292.0,
      "step": 38620
    },
    {
      "entropy": 5.8155921459197994,
      "epoch": 3.0051351877066717,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004100697260348789,
      "loss": 5.0118,
      "mean_token_accuracy": 0.1994480922818184,
      "num_tokens": 66999699.0,
      "step": 38625
    },
    {
      "entropy": 5.776117897033691,
      "epoch": 3.0055242170783893,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041004755898435254,
      "loss": 4.9244,
      "mean_token_accuracy": 0.20862591564655303,
      "num_tokens": 67008272.0,
      "step": 38630
    },
    {
      "entropy": 5.708275556564331,
      "epoch": 3.005913246450107,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004100253898846945,
      "loss": 4.9581,
      "mean_token_accuracy": 0.20586617887020112,
      "num_tokens": 67016865.0,
      "step": 38635
    },
    {
      "entropy": 5.725731945037841,
      "epoch": 3.0063022758218247,
      "grad_norm": 1.34375,
      "learning_rate": 0.000410003218736241,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21511805951595306,
      "num_tokens": 67025791.0,
      "step": 38640
    },
    {
      "entropy": 5.71353325843811,
      "epoch": 3.006691305193542,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040998104553932856,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21789900064468384,
      "num_tokens": 67034872.0,
      "step": 38645
    },
    {
      "entropy": 5.734313011169434,
      "epoch": 3.0070803345652597,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004099588702942936,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2113011360168457,
      "num_tokens": 67043350.0,
      "step": 38650
    },
    {
      "entropy": 5.7912452697753904,
      "epoch": 3.0074693639369774,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040993669300147257,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20796314328908921,
      "num_tokens": 67051593.0,
      "step": 38655
    },
    {
      "entropy": 5.849064064025879,
      "epoch": 3.0078583933086946,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040991451366120205,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20863676220178604,
      "num_tokens": 67060024.0,
      "step": 38660
    },
    {
      "entropy": 5.8618511199951175,
      "epoch": 3.0082474226804123,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004098923322738185,
      "loss": 5.0446,
      "mean_token_accuracy": 0.20807496309280396,
      "num_tokens": 67069136.0,
      "step": 38665
    },
    {
      "entropy": 5.647416591644287,
      "epoch": 3.00863645205213,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040987014883965843,
      "loss": 4.8193,
      "mean_token_accuracy": 0.22208904922008516,
      "num_tokens": 67078209.0,
      "step": 38670
    },
    {
      "entropy": 5.7008575916290285,
      "epoch": 3.0090254814238473,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040984796335905867,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21037107855081558,
      "num_tokens": 67087676.0,
      "step": 38675
    },
    {
      "entropy": 5.792073917388916,
      "epoch": 3.009414510795565,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004098257758323556,
      "loss": 4.9993,
      "mean_token_accuracy": 0.20470687448978425,
      "num_tokens": 67096730.0,
      "step": 38680
    },
    {
      "entropy": 5.899171257019043,
      "epoch": 3.0098035401672827,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004098035862598859,
      "loss": 5.0692,
      "mean_token_accuracy": 0.197377547621727,
      "num_tokens": 67106017.0,
      "step": 38685
    },
    {
      "entropy": 5.783134794235229,
      "epoch": 3.0101925695390004,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004097813946419865,
      "loss": 4.859,
      "mean_token_accuracy": 0.21705221384763718,
      "num_tokens": 67114708.0,
      "step": 38690
    },
    {
      "entropy": 5.783966064453125,
      "epoch": 3.0105815989107176,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004097592009789939,
      "loss": 4.978,
      "mean_token_accuracy": 0.20649371594190596,
      "num_tokens": 67124257.0,
      "step": 38695
    },
    {
      "entropy": 5.84283413887024,
      "epoch": 3.0109706282824353,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040973700527124494,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20660369843244553,
      "num_tokens": 67133066.0,
      "step": 38700
    },
    {
      "entropy": 5.871776437759399,
      "epoch": 3.011359657654153,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040971480751907633,
      "loss": 4.977,
      "mean_token_accuracy": 0.20864855647087097,
      "num_tokens": 67141906.0,
      "step": 38705
    },
    {
      "entropy": 5.773981380462646,
      "epoch": 3.0117486870258703,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00040969260772282494,
      "loss": 4.9535,
      "mean_token_accuracy": 0.209236678481102,
      "num_tokens": 67150533.0,
      "step": 38710
    },
    {
      "entropy": 5.7633750438690186,
      "epoch": 3.012137716397588,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004096704058828277,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20738481134176254,
      "num_tokens": 67159130.0,
      "step": 38715
    },
    {
      "entropy": 5.785203981399536,
      "epoch": 3.0125267457693057,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040964820199942135,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22140490263700485,
      "num_tokens": 67168075.0,
      "step": 38720
    },
    {
      "entropy": 5.778219890594483,
      "epoch": 3.012915775141023,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004096259960729428,
      "loss": 4.953,
      "mean_token_accuracy": 0.21326627284288407,
      "num_tokens": 67176972.0,
      "step": 38725
    },
    {
      "entropy": 5.8297162532806395,
      "epoch": 3.0133048045127406,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040960378810372906,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20281073600053787,
      "num_tokens": 67185091.0,
      "step": 38730
    },
    {
      "entropy": 5.785930156707764,
      "epoch": 3.0136938338844583,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004095815780921171,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2132229283452034,
      "num_tokens": 67193345.0,
      "step": 38735
    },
    {
      "entropy": 5.724439716339111,
      "epoch": 3.014082863256176,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004095593660384439,
      "loss": 4.887,
      "mean_token_accuracy": 0.2133231356739998,
      "num_tokens": 67201893.0,
      "step": 38740
    },
    {
      "entropy": 5.817309045791626,
      "epoch": 3.0144718926278933,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004095371519430465,
      "loss": 5.0415,
      "mean_token_accuracy": 0.20615777373313904,
      "num_tokens": 67211112.0,
      "step": 38745
    },
    {
      "entropy": 5.73439507484436,
      "epoch": 3.014860921999611,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040951493580626195,
      "loss": 4.8671,
      "mean_token_accuracy": 0.2110681101679802,
      "num_tokens": 67220078.0,
      "step": 38750
    },
    {
      "entropy": 5.751712989807129,
      "epoch": 3.0152499513713287,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004094927176284273,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2087715357542038,
      "num_tokens": 67228883.0,
      "step": 38755
    },
    {
      "entropy": 5.850134658813476,
      "epoch": 3.015638980743046,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040947049740987977,
      "loss": 5.069,
      "mean_token_accuracy": 0.20526475161314012,
      "num_tokens": 67237377.0,
      "step": 38760
    },
    {
      "entropy": 5.806525325775146,
      "epoch": 3.0160280101147636,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004094482751509565,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21030978858470917,
      "num_tokens": 67245488.0,
      "step": 38765
    },
    {
      "entropy": 5.732338809967041,
      "epoch": 3.0164170394864813,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004094260508519947,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21672680079936982,
      "num_tokens": 67253791.0,
      "step": 38770
    },
    {
      "entropy": 5.712046813964844,
      "epoch": 3.0168060688581986,
      "grad_norm": 1.453125,
      "learning_rate": 0.00040940382451333144,
      "loss": 4.882,
      "mean_token_accuracy": 0.21357179433107376,
      "num_tokens": 67262373.0,
      "step": 38775
    },
    {
      "entropy": 5.741720676422119,
      "epoch": 3.0171950982299163,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00040938159613530405,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21907440274953843,
      "num_tokens": 67270946.0,
      "step": 38780
    },
    {
      "entropy": 5.781162405014038,
      "epoch": 3.017584127601634,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004093593657182499,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21387328058481217,
      "num_tokens": 67280150.0,
      "step": 38785
    },
    {
      "entropy": 5.846461391448974,
      "epoch": 3.0179731569733517,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004093371332625062,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20985382050275803,
      "num_tokens": 67288109.0,
      "step": 38790
    },
    {
      "entropy": 5.786945247650147,
      "epoch": 3.018362186345069,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004093148987684103,
      "loss": 4.898,
      "mean_token_accuracy": 0.2122843399643898,
      "num_tokens": 67296538.0,
      "step": 38795
    },
    {
      "entropy": 5.767457580566406,
      "epoch": 3.0187512157167866,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040929266223629966,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21123732179403304,
      "num_tokens": 67306088.0,
      "step": 38800
    },
    {
      "entropy": 5.750591039657593,
      "epoch": 3.0191402450885043,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040927042366651164,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21533288359642028,
      "num_tokens": 67315101.0,
      "step": 38805
    },
    {
      "entropy": 5.76775631904602,
      "epoch": 3.0195292744602216,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040924818305938354,
      "loss": 4.863,
      "mean_token_accuracy": 0.21860644668340684,
      "num_tokens": 67323309.0,
      "step": 38810
    },
    {
      "entropy": 5.802185344696045,
      "epoch": 3.0199183038319393,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040922594041525303,
      "loss": 4.9883,
      "mean_token_accuracy": 0.20708852410316467,
      "num_tokens": 67331893.0,
      "step": 38815
    },
    {
      "entropy": 5.686135292053223,
      "epoch": 3.020307333203657,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004092036957344574,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2175297796726227,
      "num_tokens": 67339804.0,
      "step": 38820
    },
    {
      "entropy": 5.722905874252319,
      "epoch": 3.0206963625753747,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040918144901733443,
      "loss": 4.916,
      "mean_token_accuracy": 0.21310642957687378,
      "num_tokens": 67348306.0,
      "step": 38825
    },
    {
      "entropy": 5.827073621749878,
      "epoch": 3.021085391947092,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004091592002642215,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2036680445075035,
      "num_tokens": 67356601.0,
      "step": 38830
    },
    {
      "entropy": 5.711979055404663,
      "epoch": 3.0214744213188096,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004091369494754562,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2117128014564514,
      "num_tokens": 67364990.0,
      "step": 38835
    },
    {
      "entropy": 5.8275594234466555,
      "epoch": 3.0218634506905273,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004091146966513763,
      "loss": 5.0204,
      "mean_token_accuracy": 0.20446548908948897,
      "num_tokens": 67373165.0,
      "step": 38840
    },
    {
      "entropy": 5.850156497955322,
      "epoch": 3.0222524800622446,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040909244179231926,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20527103841304778,
      "num_tokens": 67381627.0,
      "step": 38845
    },
    {
      "entropy": 5.72726354598999,
      "epoch": 3.0226415094339623,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004090701848986229,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21397609263658524,
      "num_tokens": 67389928.0,
      "step": 38850
    },
    {
      "entropy": 5.730877542495728,
      "epoch": 3.02303053880568,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004090479259706247,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21561048328876495,
      "num_tokens": 67398377.0,
      "step": 38855
    },
    {
      "entropy": 5.718042230606079,
      "epoch": 3.023419568177397,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004090256650086628,
      "loss": 4.8211,
      "mean_token_accuracy": 0.2207224652171135,
      "num_tokens": 67406490.0,
      "step": 38860
    },
    {
      "entropy": 5.725705242156982,
      "epoch": 3.023808597549115,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040900340201307475,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21314817368984224,
      "num_tokens": 67415178.0,
      "step": 38865
    },
    {
      "entropy": 5.718579912185669,
      "epoch": 3.0241976269208326,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040898113698419823,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20991962254047394,
      "num_tokens": 67423770.0,
      "step": 38870
    },
    {
      "entropy": 5.812125253677368,
      "epoch": 3.0245866562925503,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004089588699223713,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20628174096345903,
      "num_tokens": 67432771.0,
      "step": 38875
    },
    {
      "entropy": 5.834966325759888,
      "epoch": 3.0249756856642676,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040893660082793167,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20966629534959794,
      "num_tokens": 67441697.0,
      "step": 38880
    },
    {
      "entropy": 5.804350090026856,
      "epoch": 3.0253647150359853,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040891432970121736,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20318198949098587,
      "num_tokens": 67450950.0,
      "step": 38885
    },
    {
      "entropy": 5.713298511505127,
      "epoch": 3.025753744407703,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040889205654256615,
      "loss": 4.8817,
      "mean_token_accuracy": 0.21540190279483795,
      "num_tokens": 67459863.0,
      "step": 38890
    },
    {
      "entropy": 5.833083152770996,
      "epoch": 3.02614277377942,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040886978135231625,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21170029044151306,
      "num_tokens": 67468829.0,
      "step": 38895
    },
    {
      "entropy": 5.7377259731292725,
      "epoch": 3.026531803151138,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040884750413080535,
      "loss": 4.8886,
      "mean_token_accuracy": 0.21775234788656234,
      "num_tokens": 67477979.0,
      "step": 38900
    },
    {
      "entropy": 5.780294036865234,
      "epoch": 3.0269208325228556,
      "grad_norm": 1.25,
      "learning_rate": 0.00040882522487837176,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2076477587223053,
      "num_tokens": 67486929.0,
      "step": 38905
    },
    {
      "entropy": 5.702690696716308,
      "epoch": 3.027309861894573,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040880294359535327,
      "loss": 4.8874,
      "mean_token_accuracy": 0.20701272785663605,
      "num_tokens": 67495434.0,
      "step": 38910
    },
    {
      "entropy": 5.782010698318482,
      "epoch": 3.0276988912662905,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040878066028208804,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2079593673348427,
      "num_tokens": 67504437.0,
      "step": 38915
    },
    {
      "entropy": 5.7263264656066895,
      "epoch": 3.0280879206380082,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040875837493891434,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21142049133777618,
      "num_tokens": 67512899.0,
      "step": 38920
    },
    {
      "entropy": 5.8580327987670895,
      "epoch": 3.028476950009726,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004087360875661701,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20422546863555907,
      "num_tokens": 67521602.0,
      "step": 38925
    },
    {
      "entropy": 5.842159461975098,
      "epoch": 3.028865979381443,
      "grad_norm": 1.375,
      "learning_rate": 0.00040871379816419376,
      "loss": 4.9696,
      "mean_token_accuracy": 0.20978191792964934,
      "num_tokens": 67530808.0,
      "step": 38930
    },
    {
      "entropy": 5.872029256820679,
      "epoch": 3.029255008753161,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004086915067333232,
      "loss": 4.935,
      "mean_token_accuracy": 0.20921196937561035,
      "num_tokens": 67539216.0,
      "step": 38935
    },
    {
      "entropy": 5.752485227584839,
      "epoch": 3.0296440381248786,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004086692132738968,
      "loss": 4.921,
      "mean_token_accuracy": 0.2113805204629898,
      "num_tokens": 67547202.0,
      "step": 38940
    },
    {
      "entropy": 5.7813129901885985,
      "epoch": 3.030033067496596,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000408646917786253,
      "loss": 5.0837,
      "mean_token_accuracy": 0.20600281804800033,
      "num_tokens": 67556427.0,
      "step": 38945
    },
    {
      "entropy": 5.832802248001099,
      "epoch": 3.0304220968683135,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004086246202707298,
      "loss": 5.009,
      "mean_token_accuracy": 0.2058857187628746,
      "num_tokens": 67564945.0,
      "step": 38950
    },
    {
      "entropy": 5.854363584518433,
      "epoch": 3.0308111262400312,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00040860232072766577,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20986813455820083,
      "num_tokens": 67574336.0,
      "step": 38955
    },
    {
      "entropy": 5.837966585159302,
      "epoch": 3.0312001556117485,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004085800191573991,
      "loss": 5.0028,
      "mean_token_accuracy": 0.2070587992668152,
      "num_tokens": 67583628.0,
      "step": 38960
    },
    {
      "entropy": 5.7908848285675045,
      "epoch": 3.031589184983466,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004085577155602683,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2085889145731926,
      "num_tokens": 67592462.0,
      "step": 38965
    },
    {
      "entropy": 5.818691062927246,
      "epoch": 3.031978214355184,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040853540993661175,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21498143523931504,
      "num_tokens": 67601377.0,
      "step": 38970
    },
    {
      "entropy": 5.725099563598633,
      "epoch": 3.0323672437269016,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040851310228676787,
      "loss": 4.846,
      "mean_token_accuracy": 0.21804891526699066,
      "num_tokens": 67610253.0,
      "step": 38975
    },
    {
      "entropy": 5.7460133075714115,
      "epoch": 3.032756273098619,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004084907926110752,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2045460596680641,
      "num_tokens": 67618092.0,
      "step": 38980
    },
    {
      "entropy": 5.754010725021362,
      "epoch": 3.0331453024703365,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004084684809098722,
      "loss": 4.9131,
      "mean_token_accuracy": 0.21501152068376542,
      "num_tokens": 67626547.0,
      "step": 38985
    },
    {
      "entropy": 5.864824676513672,
      "epoch": 3.0335343318420542,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040844616718349743,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2034089371562004,
      "num_tokens": 67635891.0,
      "step": 38990
    },
    {
      "entropy": 5.7980818271636965,
      "epoch": 3.0339233612137715,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040842385143228946,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2095649406313896,
      "num_tokens": 67645030.0,
      "step": 38995
    },
    {
      "entropy": 5.758177804946899,
      "epoch": 3.034312390585489,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040840153365658693,
      "loss": 4.981,
      "mean_token_accuracy": 0.21169472634792327,
      "num_tokens": 67652941.0,
      "step": 39000
    },
    {
      "epoch": 3.034312390585489,
      "eval_entropy": 5.4481391231180725,
      "eval_loss": 5.067856788635254,
      "eval_mean_token_accuracy": 0.21425002112454009,
      "eval_num_tokens": 67652941.0,
      "eval_runtime": 21.8126,
      "eval_samples_per_second": 1905.232,
      "eval_steps_per_second": 238.165,
      "step": 39000
    }
  ],
  "logging_steps": 5,
  "max_steps": 128520,
  "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": 1.36231096564224e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}