ind-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4668352460610776,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.92353572845459,
      "epoch": 0.00038902937171756465,
      "grad_norm": 13.375,
      "learning_rate": 2e-06,
      "loss": 14.2513,
      "mean_token_accuracy": 0.0001277139177545905,
      "num_tokens": 8755.0,
      "step": 5
    },
    {
      "entropy": 4.934064674377441,
      "epoch": 0.0007780587434351293,
      "grad_norm": 13.875,
      "learning_rate": 4.5e-06,
      "loss": 14.2685,
      "mean_token_accuracy": 0.0,
      "num_tokens": 17641.0,
      "step": 10
    },
    {
      "entropy": 4.950730228424073,
      "epoch": 0.001167088115152694,
      "grad_norm": 15.1875,
      "learning_rate": 7e-06,
      "loss": 14.1845,
      "mean_token_accuracy": 0.0,
      "num_tokens": 26293.0,
      "step": 15
    },
    {
      "entropy": 4.978141450881958,
      "epoch": 0.0015561174868702586,
      "grad_norm": 16.875,
      "learning_rate": 9.5e-06,
      "loss": 14.0758,
      "mean_token_accuracy": 0.0,
      "num_tokens": 35627.0,
      "step": 20
    },
    {
      "entropy": 5.09318265914917,
      "epoch": 0.0019451468585878235,
      "grad_norm": 20.25,
      "learning_rate": 1.2e-05,
      "loss": 13.6416,
      "mean_token_accuracy": 0.0,
      "num_tokens": 43861.0,
      "step": 25
    },
    {
      "entropy": 5.357007932662964,
      "epoch": 0.002334176230305388,
      "grad_norm": 26.375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.1478,
      "mean_token_accuracy": 0.0,
      "num_tokens": 52816.0,
      "step": 30
    },
    {
      "entropy": 6.636251544952392,
      "epoch": 0.0027232056020229526,
      "grad_norm": 25.375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 12.0885,
      "mean_token_accuracy": 0.0,
      "num_tokens": 61370.0,
      "step": 35
    },
    {
      "entropy": 9.923179340362548,
      "epoch": 0.0031122349737405172,
      "grad_norm": 3.578125,
      "learning_rate": 1.95e-05,
      "loss": 10.9048,
      "mean_token_accuracy": 0.0001186239649541676,
      "num_tokens": 69444.0,
      "step": 40
    },
    {
      "entropy": 10.702562427520752,
      "epoch": 0.003501264345458082,
      "grad_norm": 3.28125,
      "learning_rate": 2.2e-05,
      "loss": 10.6653,
      "mean_token_accuracy": 0.011188959912396967,
      "num_tokens": 77892.0,
      "step": 45
    },
    {
      "entropy": 10.734418773651123,
      "epoch": 0.003890293717175647,
      "grad_norm": 2.75,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5129,
      "mean_token_accuracy": 0.020575371105223894,
      "num_tokens": 86154.0,
      "step": 50
    },
    {
      "entropy": 10.707633876800537,
      "epoch": 0.004279323088893211,
      "grad_norm": 2.1875,
      "learning_rate": 2.7e-05,
      "loss": 10.3484,
      "mean_token_accuracy": 0.01765873283147812,
      "num_tokens": 95589.0,
      "step": 55
    },
    {
      "entropy": 10.638573741912841,
      "epoch": 0.004668352460610776,
      "grad_norm": 2.09375,
      "learning_rate": 2.95e-05,
      "loss": 10.1852,
      "mean_token_accuracy": 0.017188905738294125,
      "num_tokens": 105268.0,
      "step": 60
    },
    {
      "entropy": 10.698107624053955,
      "epoch": 0.0050573818323283405,
      "grad_norm": 1.90625,
      "learning_rate": 3.2e-05,
      "loss": 9.9755,
      "mean_token_accuracy": 0.020746274664998055,
      "num_tokens": 113347.0,
      "step": 65
    },
    {
      "entropy": 10.710992527008056,
      "epoch": 0.005446411204045905,
      "grad_norm": 1.921875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8539,
      "mean_token_accuracy": 0.029131441563367843,
      "num_tokens": 121235.0,
      "step": 70
    },
    {
      "entropy": 10.693155670166016,
      "epoch": 0.00583544057576347,
      "grad_norm": 1.734375,
      "learning_rate": 3.7e-05,
      "loss": 9.8016,
      "mean_token_accuracy": 0.03198963850736618,
      "num_tokens": 129529.0,
      "step": 75
    },
    {
      "entropy": 10.6787428855896,
      "epoch": 0.0062244699474810344,
      "grad_norm": 1.7109375,
      "learning_rate": 3.95e-05,
      "loss": 9.719,
      "mean_token_accuracy": 0.026658685877919197,
      "num_tokens": 137918.0,
      "step": 80
    },
    {
      "entropy": 10.677947998046875,
      "epoch": 0.006613499319198599,
      "grad_norm": 1.7890625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.5968,
      "mean_token_accuracy": 0.03781393375247717,
      "num_tokens": 146445.0,
      "step": 85
    },
    {
      "entropy": 10.670944595336914,
      "epoch": 0.007002528690916164,
      "grad_norm": 1.71875,
      "learning_rate": 4.45e-05,
      "loss": 9.5696,
      "mean_token_accuracy": 0.0504709430038929,
      "num_tokens": 155904.0,
      "step": 90
    },
    {
      "entropy": 10.652875709533692,
      "epoch": 0.007391558062633729,
      "grad_norm": 1.71875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.45,
      "mean_token_accuracy": 0.050913235172629356,
      "num_tokens": 164436.0,
      "step": 95
    },
    {
      "entropy": 10.624427032470702,
      "epoch": 0.007780587434351294,
      "grad_norm": 1.6328125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3318,
      "mean_token_accuracy": 0.048470248281955716,
      "num_tokens": 172985.0,
      "step": 100
    },
    {
      "entropy": 10.590660285949706,
      "epoch": 0.008169616806068859,
      "grad_norm": 1.59375,
      "learning_rate": 5.2e-05,
      "loss": 9.2226,
      "mean_token_accuracy": 0.04833147563040256,
      "num_tokens": 182084.0,
      "step": 105
    },
    {
      "entropy": 10.561868572235108,
      "epoch": 0.008558646177786422,
      "grad_norm": 1.578125,
      "learning_rate": 5.45e-05,
      "loss": 9.0932,
      "mean_token_accuracy": 0.05851382501423359,
      "num_tokens": 190742.0,
      "step": 110
    },
    {
      "entropy": 10.509655570983886,
      "epoch": 0.008947675549503988,
      "grad_norm": 1.6328125,
      "learning_rate": 5.7e-05,
      "loss": 8.9979,
      "mean_token_accuracy": 0.057733524963259696,
      "num_tokens": 199600.0,
      "step": 115
    },
    {
      "entropy": 10.445792770385742,
      "epoch": 0.009336704921221552,
      "grad_norm": 1.59375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.9087,
      "mean_token_accuracy": 0.05567027330398559,
      "num_tokens": 207883.0,
      "step": 120
    },
    {
      "entropy": 10.376938629150391,
      "epoch": 0.009725734292939117,
      "grad_norm": 1.453125,
      "learning_rate": 6.2e-05,
      "loss": 8.8037,
      "mean_token_accuracy": 0.05684689395129681,
      "num_tokens": 216216.0,
      "step": 125
    },
    {
      "entropy": 10.31500473022461,
      "epoch": 0.010114763664656681,
      "grad_norm": 1.4140625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6827,
      "mean_token_accuracy": 0.05753975734114647,
      "num_tokens": 225250.0,
      "step": 130
    },
    {
      "entropy": 10.194760513305663,
      "epoch": 0.010503793036374246,
      "grad_norm": 1.328125,
      "learning_rate": 6.7e-05,
      "loss": 8.5671,
      "mean_token_accuracy": 0.054637744277715686,
      "num_tokens": 233748.0,
      "step": 135
    },
    {
      "entropy": 10.12427225112915,
      "epoch": 0.01089282240809181,
      "grad_norm": 1.1796875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.4612,
      "mean_token_accuracy": 0.0579361479729414,
      "num_tokens": 242629.0,
      "step": 140
    },
    {
      "entropy": 9.959631061553955,
      "epoch": 0.011281851779809376,
      "grad_norm": 1.421875,
      "learning_rate": 7.2e-05,
      "loss": 8.3515,
      "mean_token_accuracy": 0.054840120300650594,
      "num_tokens": 251458.0,
      "step": 145
    },
    {
      "entropy": 9.83036823272705,
      "epoch": 0.01167088115152694,
      "grad_norm": 1.1796875,
      "learning_rate": 7.45e-05,
      "loss": 8.2394,
      "mean_token_accuracy": 0.06350252479314804,
      "num_tokens": 260557.0,
      "step": 150
    },
    {
      "entropy": 9.662144470214844,
      "epoch": 0.012059910523244505,
      "grad_norm": 1.0703125,
      "learning_rate": 7.7e-05,
      "loss": 8.1841,
      "mean_token_accuracy": 0.061289533600211145,
      "num_tokens": 269262.0,
      "step": 155
    },
    {
      "entropy": 9.52525873184204,
      "epoch": 0.012448939894962069,
      "grad_norm": 1.0234375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.0687,
      "mean_token_accuracy": 0.06112904436886311,
      "num_tokens": 277804.0,
      "step": 160
    },
    {
      "entropy": 9.351448822021485,
      "epoch": 0.012837969266679634,
      "grad_norm": 1.1953125,
      "learning_rate": 8.2e-05,
      "loss": 7.9878,
      "mean_token_accuracy": 0.06238803043961525,
      "num_tokens": 287450.0,
      "step": 165
    },
    {
      "entropy": 9.164531326293945,
      "epoch": 0.013226998638397198,
      "grad_norm": 0.99609375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.947,
      "mean_token_accuracy": 0.05899545960128307,
      "num_tokens": 296497.0,
      "step": 170
    },
    {
      "entropy": 8.984291362762452,
      "epoch": 0.013616028010114764,
      "grad_norm": 0.83984375,
      "learning_rate": 8.7e-05,
      "loss": 7.8233,
      "mean_token_accuracy": 0.06463964357972145,
      "num_tokens": 305414.0,
      "step": 175
    },
    {
      "entropy": 8.76324167251587,
      "epoch": 0.014005057381832327,
      "grad_norm": 0.87109375,
      "learning_rate": 8.95e-05,
      "loss": 7.8054,
      "mean_token_accuracy": 0.05876799784600735,
      "num_tokens": 313639.0,
      "step": 180
    },
    {
      "entropy": 8.647741985321044,
      "epoch": 0.014394086753549893,
      "grad_norm": 0.73828125,
      "learning_rate": 9.2e-05,
      "loss": 7.7893,
      "mean_token_accuracy": 0.06448987349867821,
      "num_tokens": 322855.0,
      "step": 185
    },
    {
      "entropy": 8.506945037841797,
      "epoch": 0.014783116125267459,
      "grad_norm": 0.7265625,
      "learning_rate": 9.45e-05,
      "loss": 7.8651,
      "mean_token_accuracy": 0.06298348568379879,
      "num_tokens": 332071.0,
      "step": 190
    },
    {
      "entropy": 8.408401393890381,
      "epoch": 0.015172145496985022,
      "grad_norm": 0.8046875,
      "learning_rate": 9.7e-05,
      "loss": 7.668,
      "mean_token_accuracy": 0.06562523804605007,
      "num_tokens": 339990.0,
      "step": 195
    },
    {
      "entropy": 8.309197616577148,
      "epoch": 0.015561174868702588,
      "grad_norm": 0.83984375,
      "learning_rate": 9.95e-05,
      "loss": 7.7209,
      "mean_token_accuracy": 0.0656034205108881,
      "num_tokens": 348581.0,
      "step": 200
    },
    {
      "entropy": 8.265014839172363,
      "epoch": 0.01595020424042015,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000102,
      "loss": 7.709,
      "mean_token_accuracy": 0.06688988246023655,
      "num_tokens": 357032.0,
      "step": 205
    },
    {
      "entropy": 8.248595333099365,
      "epoch": 0.016339233612137717,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.6737,
      "mean_token_accuracy": 0.06308167651295662,
      "num_tokens": 366562.0,
      "step": 210
    },
    {
      "entropy": 8.197728157043457,
      "epoch": 0.016728262983855283,
      "grad_norm": 0.828125,
      "learning_rate": 0.000107,
      "loss": 7.6647,
      "mean_token_accuracy": 0.06641935594379902,
      "num_tokens": 374555.0,
      "step": 215
    },
    {
      "entropy": 8.081507778167724,
      "epoch": 0.017117292355572845,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0001095,
      "loss": 7.702,
      "mean_token_accuracy": 0.060842974483966826,
      "num_tokens": 383583.0,
      "step": 220
    },
    {
      "entropy": 8.141448211669921,
      "epoch": 0.01750632172729041,
      "grad_norm": 0.6484375,
      "learning_rate": 0.000112,
      "loss": 7.719,
      "mean_token_accuracy": 0.06294353380799293,
      "num_tokens": 392932.0,
      "step": 225
    },
    {
      "entropy": 8.129082298278808,
      "epoch": 0.017895351099007976,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0001145,
      "loss": 7.572,
      "mean_token_accuracy": 0.06622617989778519,
      "num_tokens": 401653.0,
      "step": 230
    },
    {
      "entropy": 8.068495750427246,
      "epoch": 0.01828438047072554,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.6611,
      "mean_token_accuracy": 0.06250618062913418,
      "num_tokens": 410432.0,
      "step": 235
    },
    {
      "entropy": 8.076781940460204,
      "epoch": 0.018673409842443103,
      "grad_norm": 0.71875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.6261,
      "mean_token_accuracy": 0.06318732835352421,
      "num_tokens": 419380.0,
      "step": 240
    },
    {
      "entropy": 8.065077304840088,
      "epoch": 0.01906243921416067,
      "grad_norm": 1.0,
      "learning_rate": 0.000122,
      "loss": 7.5564,
      "mean_token_accuracy": 0.0642390213906765,
      "num_tokens": 427682.0,
      "step": 245
    },
    {
      "entropy": 8.04588441848755,
      "epoch": 0.019451468585878234,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0001245,
      "loss": 7.6214,
      "mean_token_accuracy": 0.062338390946388246,
      "num_tokens": 437708.0,
      "step": 250
    },
    {
      "entropy": 8.039131164550781,
      "epoch": 0.0198404979575958,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000127,
      "loss": 7.6017,
      "mean_token_accuracy": 0.067028983309865,
      "num_tokens": 446480.0,
      "step": 255
    },
    {
      "entropy": 8.057591438293457,
      "epoch": 0.020229527329313362,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0001295,
      "loss": 7.6213,
      "mean_token_accuracy": 0.06589528843760491,
      "num_tokens": 455273.0,
      "step": 260
    },
    {
      "entropy": 8.022430849075317,
      "epoch": 0.020618556701030927,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000132,
      "loss": 7.5587,
      "mean_token_accuracy": 0.06589296795427799,
      "num_tokens": 464306.0,
      "step": 265
    },
    {
      "entropy": 8.003626775741576,
      "epoch": 0.021007586072748493,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.5562,
      "mean_token_accuracy": 0.06347908154129982,
      "num_tokens": 472812.0,
      "step": 270
    },
    {
      "entropy": 7.978896903991699,
      "epoch": 0.02139661544446606,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.5405,
      "mean_token_accuracy": 0.06544829607009887,
      "num_tokens": 481326.0,
      "step": 275
    },
    {
      "entropy": 7.993354463577271,
      "epoch": 0.02178564481618362,
      "grad_norm": 0.875,
      "learning_rate": 0.0001395,
      "loss": 7.4823,
      "mean_token_accuracy": 0.06848083920776844,
      "num_tokens": 489147.0,
      "step": 280
    },
    {
      "entropy": 7.981906986236572,
      "epoch": 0.022174674187901186,
      "grad_norm": 0.84375,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.4842,
      "mean_token_accuracy": 0.07221479900181293,
      "num_tokens": 497455.0,
      "step": 285
    },
    {
      "entropy": 7.967179822921753,
      "epoch": 0.02256370355961875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001445,
      "loss": 7.4328,
      "mean_token_accuracy": 0.0648717537522316,
      "num_tokens": 506735.0,
      "step": 290
    },
    {
      "entropy": 7.938949966430664,
      "epoch": 0.022952732931336317,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000147,
      "loss": 7.489,
      "mean_token_accuracy": 0.06714184954762459,
      "num_tokens": 515976.0,
      "step": 295
    },
    {
      "entropy": 7.8373462677001955,
      "epoch": 0.02334176230305388,
      "grad_norm": 0.84375,
      "learning_rate": 0.0001495,
      "loss": 7.4997,
      "mean_token_accuracy": 0.06925978623330593,
      "num_tokens": 524986.0,
      "step": 300
    },
    {
      "entropy": 7.921111154556274,
      "epoch": 0.023730791674771445,
      "grad_norm": 0.984375,
      "learning_rate": 0.000152,
      "loss": 7.4968,
      "mean_token_accuracy": 0.07236363515257835,
      "num_tokens": 533146.0,
      "step": 305
    },
    {
      "entropy": 7.893795490264893,
      "epoch": 0.02411982104648901,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.4158,
      "mean_token_accuracy": 0.06991111040115357,
      "num_tokens": 541704.0,
      "step": 310
    },
    {
      "entropy": 7.880386638641357,
      "epoch": 0.024508850418206576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000157,
      "loss": 7.3869,
      "mean_token_accuracy": 0.0700387105345726,
      "num_tokens": 550088.0,
      "step": 315
    },
    {
      "entropy": 7.82145209312439,
      "epoch": 0.024897879789924138,
      "grad_norm": 1.0,
      "learning_rate": 0.0001595,
      "loss": 7.3733,
      "mean_token_accuracy": 0.07134931497275829,
      "num_tokens": 558763.0,
      "step": 320
    },
    {
      "entropy": 7.873892116546631,
      "epoch": 0.025286909161641703,
      "grad_norm": 0.6953125,
      "learning_rate": 0.000162,
      "loss": 7.4578,
      "mean_token_accuracy": 0.06747227981686592,
      "num_tokens": 567538.0,
      "step": 325
    },
    {
      "entropy": 7.873815393447876,
      "epoch": 0.02567593853335927,
      "grad_norm": 1.03125,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.4837,
      "mean_token_accuracy": 0.0674517210572958,
      "num_tokens": 577626.0,
      "step": 330
    },
    {
      "entropy": 7.768797588348389,
      "epoch": 0.026064967905076834,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.4164,
      "mean_token_accuracy": 0.06899687238037586,
      "num_tokens": 587163.0,
      "step": 335
    },
    {
      "entropy": 7.8342077255249025,
      "epoch": 0.026453997276794396,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.3806,
      "mean_token_accuracy": 0.07264294996857643,
      "num_tokens": 595896.0,
      "step": 340
    },
    {
      "entropy": 7.910291004180908,
      "epoch": 0.026843026648511962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.409,
      "mean_token_accuracy": 0.07169737853109837,
      "num_tokens": 603815.0,
      "step": 345
    },
    {
      "entropy": 7.82722544670105,
      "epoch": 0.027232056020229527,
      "grad_norm": 1.671875,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.4483,
      "mean_token_accuracy": 0.07086139768362046,
      "num_tokens": 612208.0,
      "step": 350
    },
    {
      "entropy": 7.742429780960083,
      "epoch": 0.027621085391947093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000177,
      "loss": 7.4119,
      "mean_token_accuracy": 0.07163332626223565,
      "num_tokens": 621840.0,
      "step": 355
    },
    {
      "entropy": 7.883988571166992,
      "epoch": 0.028010114763664655,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0001795,
      "loss": 7.4061,
      "mean_token_accuracy": 0.07186231873929501,
      "num_tokens": 630770.0,
      "step": 360
    },
    {
      "entropy": 7.773575687408448,
      "epoch": 0.02839914413538222,
      "grad_norm": 1.234375,
      "learning_rate": 0.000182,
      "loss": 7.4064,
      "mean_token_accuracy": 0.07286142185330391,
      "num_tokens": 640313.0,
      "step": 365
    },
    {
      "entropy": 7.860558700561524,
      "epoch": 0.028788173507099786,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001845,
      "loss": 7.4376,
      "mean_token_accuracy": 0.07616207674145699,
      "num_tokens": 648594.0,
      "step": 370
    },
    {
      "entropy": 7.8611939430236815,
      "epoch": 0.02917720287881735,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000187,
      "loss": 7.4204,
      "mean_token_accuracy": 0.07530638352036476,
      "num_tokens": 656478.0,
      "step": 375
    },
    {
      "entropy": 7.69996657371521,
      "epoch": 0.029566232250534917,
      "grad_norm": 1.0,
      "learning_rate": 0.0001895,
      "loss": 7.3901,
      "mean_token_accuracy": 0.07715101987123489,
      "num_tokens": 665405.0,
      "step": 380
    },
    {
      "entropy": 7.884545755386353,
      "epoch": 0.02995526162225248,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000192,
      "loss": 7.4072,
      "mean_token_accuracy": 0.07346850857138634,
      "num_tokens": 674301.0,
      "step": 385
    },
    {
      "entropy": 7.79351954460144,
      "epoch": 0.030344290993970045,
      "grad_norm": 1.21875,
      "learning_rate": 0.0001945,
      "loss": 7.4109,
      "mean_token_accuracy": 0.08006449043750763,
      "num_tokens": 683159.0,
      "step": 390
    },
    {
      "entropy": 7.635227489471435,
      "epoch": 0.03073332036568761,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.253,
      "mean_token_accuracy": 0.08206554427742958,
      "num_tokens": 690992.0,
      "step": 395
    },
    {
      "entropy": 7.689710521697998,
      "epoch": 0.031122349737405176,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.3351,
      "mean_token_accuracy": 0.08058228120207786,
      "num_tokens": 699934.0,
      "step": 400
    },
    {
      "entropy": 7.84839277267456,
      "epoch": 0.03151137910912274,
      "grad_norm": 0.96875,
      "learning_rate": 0.000202,
      "loss": 7.3736,
      "mean_token_accuracy": 0.07575208507478237,
      "num_tokens": 708844.0,
      "step": 405
    },
    {
      "entropy": 7.728482341766357,
      "epoch": 0.0319004084808403,
      "grad_norm": 0.90625,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.4388,
      "mean_token_accuracy": 0.07045924849808216,
      "num_tokens": 718148.0,
      "step": 410
    },
    {
      "entropy": 7.762258625030517,
      "epoch": 0.03228943785255787,
      "grad_norm": 1.09375,
      "learning_rate": 0.000207,
      "loss": 7.3623,
      "mean_token_accuracy": 0.0722539782524109,
      "num_tokens": 726859.0,
      "step": 415
    },
    {
      "entropy": 7.745457887649536,
      "epoch": 0.032678467224275434,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0002095,
      "loss": 7.3553,
      "mean_token_accuracy": 0.07971226572990417,
      "num_tokens": 735045.0,
      "step": 420
    },
    {
      "entropy": 7.745830869674682,
      "epoch": 0.033067496595993,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000212,
      "loss": 7.2402,
      "mean_token_accuracy": 0.08178875669836998,
      "num_tokens": 743538.0,
      "step": 425
    },
    {
      "entropy": 7.6178747653961185,
      "epoch": 0.033456525967710565,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002145,
      "loss": 7.2988,
      "mean_token_accuracy": 0.0807783305644989,
      "num_tokens": 752518.0,
      "step": 430
    },
    {
      "entropy": 7.715822649002075,
      "epoch": 0.033845555339428124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.3213,
      "mean_token_accuracy": 0.07773758620023727,
      "num_tokens": 760354.0,
      "step": 435
    },
    {
      "entropy": 7.711114501953125,
      "epoch": 0.03423458471114569,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002195,
      "loss": 7.3633,
      "mean_token_accuracy": 0.07912566363811493,
      "num_tokens": 769522.0,
      "step": 440
    },
    {
      "entropy": 7.787883377075195,
      "epoch": 0.034623614082863255,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000222,
      "loss": 7.3054,
      "mean_token_accuracy": 0.08274181857705117,
      "num_tokens": 778548.0,
      "step": 445
    },
    {
      "entropy": 7.620460271835327,
      "epoch": 0.03501264345458082,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002245,
      "loss": 7.318,
      "mean_token_accuracy": 0.07461104467511177,
      "num_tokens": 787666.0,
      "step": 450
    },
    {
      "entropy": 7.71654577255249,
      "epoch": 0.035401672826298386,
      "grad_norm": 1.15625,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.2207,
      "mean_token_accuracy": 0.07886159271001816,
      "num_tokens": 796550.0,
      "step": 455
    },
    {
      "entropy": 7.6406464099884035,
      "epoch": 0.03579070219801595,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.275,
      "mean_token_accuracy": 0.083228849619627,
      "num_tokens": 804549.0,
      "step": 460
    },
    {
      "entropy": 7.61014609336853,
      "epoch": 0.03617973156973352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.2721,
      "mean_token_accuracy": 0.08080526292324067,
      "num_tokens": 813328.0,
      "step": 465
    },
    {
      "entropy": 7.666672468185425,
      "epoch": 0.03656876094145108,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.2935,
      "mean_token_accuracy": 0.08522161021828652,
      "num_tokens": 821733.0,
      "step": 470
    },
    {
      "entropy": 7.678701782226563,
      "epoch": 0.03695779031316864,
      "grad_norm": 0.890625,
      "learning_rate": 0.000237,
      "loss": 7.2352,
      "mean_token_accuracy": 0.08546577319502831,
      "num_tokens": 830243.0,
      "step": 475
    },
    {
      "entropy": 7.726054334640503,
      "epoch": 0.03734681968488621,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002395,
      "loss": 7.2069,
      "mean_token_accuracy": 0.08190477639436722,
      "num_tokens": 839059.0,
      "step": 480
    },
    {
      "entropy": 7.506451272964478,
      "epoch": 0.03773584905660377,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000242,
      "loss": 7.1162,
      "mean_token_accuracy": 0.08447866290807723,
      "num_tokens": 846932.0,
      "step": 485
    },
    {
      "entropy": 7.670350217819214,
      "epoch": 0.03812487842832134,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002445,
      "loss": 7.2847,
      "mean_token_accuracy": 0.07881609201431275,
      "num_tokens": 855333.0,
      "step": 490
    },
    {
      "entropy": 7.547208404541015,
      "epoch": 0.0385139078000389,
      "grad_norm": 1.1875,
      "learning_rate": 0.000247,
      "loss": 7.0816,
      "mean_token_accuracy": 0.08268252983689309,
      "num_tokens": 863571.0,
      "step": 495
    },
    {
      "entropy": 7.472938203811646,
      "epoch": 0.03890293717175647,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002495,
      "loss": 7.1508,
      "mean_token_accuracy": 0.0897403359413147,
      "num_tokens": 872145.0,
      "step": 500
    },
    {
      "entropy": 7.586430358886719,
      "epoch": 0.039291966543474034,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000252,
      "loss": 7.1937,
      "mean_token_accuracy": 0.08515456318855286,
      "num_tokens": 880634.0,
      "step": 505
    },
    {
      "entropy": 7.575957155227661,
      "epoch": 0.0396809959151916,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002545,
      "loss": 7.1256,
      "mean_token_accuracy": 0.08602449670433998,
      "num_tokens": 889178.0,
      "step": 510
    },
    {
      "entropy": 7.5376099109649655,
      "epoch": 0.04007002528690916,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000257,
      "loss": 7.1261,
      "mean_token_accuracy": 0.08836784139275551,
      "num_tokens": 898255.0,
      "step": 515
    },
    {
      "entropy": 7.452177858352661,
      "epoch": 0.040459054658626724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002595,
      "loss": 7.1758,
      "mean_token_accuracy": 0.08208275288343429,
      "num_tokens": 907369.0,
      "step": 520
    },
    {
      "entropy": 7.568976831436157,
      "epoch": 0.04084808403034429,
      "grad_norm": 1.140625,
      "learning_rate": 0.000262,
      "loss": 7.0939,
      "mean_token_accuracy": 0.08972274139523506,
      "num_tokens": 915946.0,
      "step": 525
    },
    {
      "entropy": 7.516032934188843,
      "epoch": 0.041237113402061855,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.1498,
      "mean_token_accuracy": 0.084378083050251,
      "num_tokens": 925579.0,
      "step": 530
    },
    {
      "entropy": 7.533543968200684,
      "epoch": 0.04162614277377942,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.1322,
      "mean_token_accuracy": 0.07946365103125572,
      "num_tokens": 934732.0,
      "step": 535
    },
    {
      "entropy": 7.449686288833618,
      "epoch": 0.042015172145496986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.1725,
      "mean_token_accuracy": 0.08084325268864631,
      "num_tokens": 943371.0,
      "step": 540
    },
    {
      "entropy": 7.500879001617432,
      "epoch": 0.04240420151721455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.0484,
      "mean_token_accuracy": 0.0895972803235054,
      "num_tokens": 951977.0,
      "step": 545
    },
    {
      "entropy": 7.4533003807067875,
      "epoch": 0.04279323088893212,
      "grad_norm": 1.125,
      "learning_rate": 0.0002745,
      "loss": 7.0355,
      "mean_token_accuracy": 0.08960011452436448,
      "num_tokens": 960474.0,
      "step": 550
    },
    {
      "entropy": 7.598705387115478,
      "epoch": 0.04318226026064968,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000277,
      "loss": 7.1391,
      "mean_token_accuracy": 0.0912420243024826,
      "num_tokens": 969823.0,
      "step": 555
    },
    {
      "entropy": 7.485738754272461,
      "epoch": 0.04357128963236724,
      "grad_norm": 1.25,
      "learning_rate": 0.0002795,
      "loss": 7.0609,
      "mean_token_accuracy": 0.09367068633437156,
      "num_tokens": 977417.0,
      "step": 560
    },
    {
      "entropy": 7.455179023742676,
      "epoch": 0.04396031900408481,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.0682,
      "mean_token_accuracy": 0.0958904579281807,
      "num_tokens": 985519.0,
      "step": 565
    },
    {
      "entropy": 7.482391786575318,
      "epoch": 0.04434934837580237,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002845,
      "loss": 7.0641,
      "mean_token_accuracy": 0.08767548501491547,
      "num_tokens": 994136.0,
      "step": 570
    },
    {
      "entropy": 7.325763988494873,
      "epoch": 0.04473837774751994,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000287,
      "loss": 7.0354,
      "mean_token_accuracy": 0.08815001025795936,
      "num_tokens": 1002896.0,
      "step": 575
    },
    {
      "entropy": 7.49589991569519,
      "epoch": 0.0451274071192375,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002895,
      "loss": 7.1115,
      "mean_token_accuracy": 0.08954815194010735,
      "num_tokens": 1011032.0,
      "step": 580
    },
    {
      "entropy": 7.365485429763794,
      "epoch": 0.04551643649095507,
      "grad_norm": 1.25,
      "learning_rate": 0.000292,
      "loss": 7.0774,
      "mean_token_accuracy": 0.08928431719541549,
      "num_tokens": 1019901.0,
      "step": 585
    },
    {
      "entropy": 7.4467432498931885,
      "epoch": 0.045905465862672634,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002945,
      "loss": 7.0539,
      "mean_token_accuracy": 0.0877431645989418,
      "num_tokens": 1029049.0,
      "step": 590
    },
    {
      "entropy": 7.367093896865844,
      "epoch": 0.0462944952343902,
      "grad_norm": 0.90625,
      "learning_rate": 0.000297,
      "loss": 7.0025,
      "mean_token_accuracy": 0.09999658241868019,
      "num_tokens": 1037580.0,
      "step": 595
    },
    {
      "entropy": 7.53356819152832,
      "epoch": 0.04668352460610776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002995,
      "loss": 7.0789,
      "mean_token_accuracy": 0.09073482006788254,
      "num_tokens": 1045804.0,
      "step": 600
    },
    {
      "entropy": 7.356887912750244,
      "epoch": 0.047072553977825324,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000302,
      "loss": 7.0293,
      "mean_token_accuracy": 0.08818676322698593,
      "num_tokens": 1054719.0,
      "step": 605
    },
    {
      "entropy": 7.446684646606445,
      "epoch": 0.04746158334954289,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003045,
      "loss": 7.013,
      "mean_token_accuracy": 0.08933027312159539,
      "num_tokens": 1063373.0,
      "step": 610
    },
    {
      "entropy": 7.305212068557739,
      "epoch": 0.047850612721260455,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000307,
      "loss": 6.9932,
      "mean_token_accuracy": 0.09075823351740837,
      "num_tokens": 1073021.0,
      "step": 615
    },
    {
      "entropy": 7.409698390960694,
      "epoch": 0.04823964209297802,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003095,
      "loss": 7.0495,
      "mean_token_accuracy": 0.08623504415154457,
      "num_tokens": 1081547.0,
      "step": 620
    },
    {
      "entropy": 7.397112703323364,
      "epoch": 0.048628671464695586,
      "grad_norm": 1.25,
      "learning_rate": 0.000312,
      "loss": 7.0104,
      "mean_token_accuracy": 0.08981771692633629,
      "num_tokens": 1090695.0,
      "step": 625
    },
    {
      "entropy": 7.31109766960144,
      "epoch": 0.04901770083641315,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003145,
      "loss": 6.9118,
      "mean_token_accuracy": 0.0954255647957325,
      "num_tokens": 1099563.0,
      "step": 630
    },
    {
      "entropy": 7.338528490066528,
      "epoch": 0.04940673020813072,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000317,
      "loss": 6.9835,
      "mean_token_accuracy": 0.09025626108050347,
      "num_tokens": 1109021.0,
      "step": 635
    },
    {
      "entropy": 7.372527933120727,
      "epoch": 0.049795759579848276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003195,
      "loss": 6.9745,
      "mean_token_accuracy": 0.09381573870778084,
      "num_tokens": 1118281.0,
      "step": 640
    },
    {
      "entropy": 7.344166088104248,
      "epoch": 0.05018478895156584,
      "grad_norm": 1.046875,
      "learning_rate": 0.000322,
      "loss": 7.0264,
      "mean_token_accuracy": 0.08852934762835503,
      "num_tokens": 1127513.0,
      "step": 645
    },
    {
      "entropy": 7.236174535751343,
      "epoch": 0.05057381832328341,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.8914,
      "mean_token_accuracy": 0.09319968894124031,
      "num_tokens": 1136140.0,
      "step": 650
    },
    {
      "entropy": 7.3045234203338625,
      "epoch": 0.05096284769500097,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.9159,
      "mean_token_accuracy": 0.09398315250873565,
      "num_tokens": 1144297.0,
      "step": 655
    },
    {
      "entropy": 7.339838361740112,
      "epoch": 0.05135187706671854,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.0013,
      "mean_token_accuracy": 0.0909078523516655,
      "num_tokens": 1153639.0,
      "step": 660
    },
    {
      "entropy": 7.2977570533752445,
      "epoch": 0.0517409064384361,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.9892,
      "mean_token_accuracy": 0.09121011421084405,
      "num_tokens": 1162825.0,
      "step": 665
    },
    {
      "entropy": 7.272716379165649,
      "epoch": 0.05212993581015367,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.0,
      "mean_token_accuracy": 0.0927580676972866,
      "num_tokens": 1171887.0,
      "step": 670
    },
    {
      "entropy": 7.289681434631348,
      "epoch": 0.052518965181871234,
      "grad_norm": 1.265625,
      "learning_rate": 0.000337,
      "loss": 6.9128,
      "mean_token_accuracy": 0.09554491266608238,
      "num_tokens": 1179963.0,
      "step": 675
    },
    {
      "entropy": 7.24162278175354,
      "epoch": 0.05290799455358879,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003395,
      "loss": 6.8815,
      "mean_token_accuracy": 0.09879456236958503,
      "num_tokens": 1188408.0,
      "step": 680
    },
    {
      "entropy": 7.2835136413574215,
      "epoch": 0.05329702392530636,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000342,
      "loss": 6.9133,
      "mean_token_accuracy": 0.0909463919699192,
      "num_tokens": 1197751.0,
      "step": 685
    },
    {
      "entropy": 7.258634185791015,
      "epoch": 0.053686053297023924,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.9544,
      "mean_token_accuracy": 0.09061150550842285,
      "num_tokens": 1207101.0,
      "step": 690
    },
    {
      "entropy": 7.253326082229615,
      "epoch": 0.05407508266874149,
      "grad_norm": 1.078125,
      "learning_rate": 0.000347,
      "loss": 6.9457,
      "mean_token_accuracy": 0.09488306194543839,
      "num_tokens": 1216395.0,
      "step": 695
    },
    {
      "entropy": 7.289101219177246,
      "epoch": 0.054464112040459055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003495,
      "loss": 6.9216,
      "mean_token_accuracy": 0.09484536051750184,
      "num_tokens": 1225496.0,
      "step": 700
    },
    {
      "entropy": 7.246467351913452,
      "epoch": 0.05485314141217662,
      "grad_norm": 1.0625,
      "learning_rate": 0.000352,
      "loss": 6.9826,
      "mean_token_accuracy": 0.09187191799283027,
      "num_tokens": 1233934.0,
      "step": 705
    },
    {
      "entropy": 7.28291802406311,
      "epoch": 0.055242170783894186,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003545,
      "loss": 6.8526,
      "mean_token_accuracy": 0.10444689691066741,
      "num_tokens": 1242165.0,
      "step": 710
    },
    {
      "entropy": 7.235671281814575,
      "epoch": 0.05563120015561175,
      "grad_norm": 1.421875,
      "learning_rate": 0.000357,
      "loss": 6.9106,
      "mean_token_accuracy": 0.09974046424031258,
      "num_tokens": 1250021.0,
      "step": 715
    },
    {
      "entropy": 7.2171196937561035,
      "epoch": 0.05602022952732931,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003595,
      "loss": 6.9197,
      "mean_token_accuracy": 0.09527276083827019,
      "num_tokens": 1258397.0,
      "step": 720
    },
    {
      "entropy": 7.222966718673706,
      "epoch": 0.056409258899046875,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000362,
      "loss": 6.9115,
      "mean_token_accuracy": 0.09537832140922546,
      "num_tokens": 1266764.0,
      "step": 725
    },
    {
      "entropy": 7.289536237716675,
      "epoch": 0.05679828827076444,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003645,
      "loss": 6.9214,
      "mean_token_accuracy": 0.09203095957636834,
      "num_tokens": 1275643.0,
      "step": 730
    },
    {
      "entropy": 7.227653741836548,
      "epoch": 0.057187317642482006,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000367,
      "loss": 6.9059,
      "mean_token_accuracy": 0.0964889533817768,
      "num_tokens": 1284379.0,
      "step": 735
    },
    {
      "entropy": 7.118802309036255,
      "epoch": 0.05757634701419957,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003695,
      "loss": 6.7643,
      "mean_token_accuracy": 0.10291910842061043,
      "num_tokens": 1293115.0,
      "step": 740
    },
    {
      "entropy": 7.243952751159668,
      "epoch": 0.05796537638591714,
      "grad_norm": 1.1875,
      "learning_rate": 0.000372,
      "loss": 6.9463,
      "mean_token_accuracy": 0.09045971632003784,
      "num_tokens": 1301583.0,
      "step": 745
    },
    {
      "entropy": 7.1956870555877686,
      "epoch": 0.0583544057576347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003745,
      "loss": 6.8552,
      "mean_token_accuracy": 0.09680608436465263,
      "num_tokens": 1310721.0,
      "step": 750
    },
    {
      "entropy": 7.304637289047241,
      "epoch": 0.05874343512935227,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000377,
      "loss": 6.9913,
      "mean_token_accuracy": 0.0911169946193695,
      "num_tokens": 1319056.0,
      "step": 755
    },
    {
      "entropy": 7.180689954757691,
      "epoch": 0.059132464501069834,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003795,
      "loss": 6.8737,
      "mean_token_accuracy": 0.09462213069200516,
      "num_tokens": 1327259.0,
      "step": 760
    },
    {
      "entropy": 7.10323920249939,
      "epoch": 0.05952149387278739,
      "grad_norm": 1.0625,
      "learning_rate": 0.000382,
      "loss": 6.7893,
      "mean_token_accuracy": 0.09687989726662635,
      "num_tokens": 1335906.0,
      "step": 765
    },
    {
      "entropy": 7.191504859924317,
      "epoch": 0.05991052324450496,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003845,
      "loss": 6.832,
      "mean_token_accuracy": 0.09174941778182984,
      "num_tokens": 1344743.0,
      "step": 770
    },
    {
      "entropy": 7.170624208450318,
      "epoch": 0.060299552616222524,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.8469,
      "mean_token_accuracy": 0.09922058656811714,
      "num_tokens": 1353211.0,
      "step": 775
    },
    {
      "entropy": 7.190778684616089,
      "epoch": 0.06068858198794009,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.954,
      "mean_token_accuracy": 0.0928943283855915,
      "num_tokens": 1361849.0,
      "step": 780
    },
    {
      "entropy": 7.262214231491089,
      "epoch": 0.061077611359657655,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.9363,
      "mean_token_accuracy": 0.09803319126367568,
      "num_tokens": 1369804.0,
      "step": 785
    },
    {
      "entropy": 7.06411600112915,
      "epoch": 0.06146664073137522,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.7328,
      "mean_token_accuracy": 0.09864451736211777,
      "num_tokens": 1378478.0,
      "step": 790
    },
    {
      "entropy": 7.197555208206177,
      "epoch": 0.061855670103092786,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.8123,
      "mean_token_accuracy": 0.10194888114929199,
      "num_tokens": 1387225.0,
      "step": 795
    },
    {
      "entropy": 7.082095813751221,
      "epoch": 0.06224469947481035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003995,
      "loss": 6.7738,
      "mean_token_accuracy": 0.10356510728597641,
      "num_tokens": 1395338.0,
      "step": 800
    },
    {
      "entropy": 7.111552381515503,
      "epoch": 0.06263372884652792,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000402,
      "loss": 6.7695,
      "mean_token_accuracy": 0.10290253907442093,
      "num_tokens": 1403794.0,
      "step": 805
    },
    {
      "entropy": 7.024047470092773,
      "epoch": 0.06302275821824548,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004045,
      "loss": 6.8432,
      "mean_token_accuracy": 0.09563084170222283,
      "num_tokens": 1412522.0,
      "step": 810
    },
    {
      "entropy": 7.1450153350830075,
      "epoch": 0.06341178758996305,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.8187,
      "mean_token_accuracy": 0.09947089925408363,
      "num_tokens": 1421347.0,
      "step": 815
    },
    {
      "entropy": 7.098122262954712,
      "epoch": 0.0638008169616806,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004095,
      "loss": 6.8049,
      "mean_token_accuracy": 0.10245971083641052,
      "num_tokens": 1429788.0,
      "step": 820
    },
    {
      "entropy": 7.155891752243042,
      "epoch": 0.06418984633339817,
      "grad_norm": 1.171875,
      "learning_rate": 0.000412,
      "loss": 6.8032,
      "mean_token_accuracy": 0.09958769977092743,
      "num_tokens": 1437442.0,
      "step": 825
    },
    {
      "entropy": 7.052066421508789,
      "epoch": 0.06457887570511574,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004145,
      "loss": 6.7311,
      "mean_token_accuracy": 0.1029736690223217,
      "num_tokens": 1445604.0,
      "step": 830
    },
    {
      "entropy": 7.043609714508056,
      "epoch": 0.0649679050768333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000417,
      "loss": 6.7971,
      "mean_token_accuracy": 0.10040144473314286,
      "num_tokens": 1454342.0,
      "step": 835
    },
    {
      "entropy": 7.063785266876221,
      "epoch": 0.06535693444855087,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004195,
      "loss": 6.783,
      "mean_token_accuracy": 0.10244178175926208,
      "num_tokens": 1463189.0,
      "step": 840
    },
    {
      "entropy": 7.155404758453369,
      "epoch": 0.06574596382026843,
      "grad_norm": 1.171875,
      "learning_rate": 0.000422,
      "loss": 6.8392,
      "mean_token_accuracy": 0.10035181567072868,
      "num_tokens": 1471608.0,
      "step": 845
    },
    {
      "entropy": 7.050847911834717,
      "epoch": 0.066134993191986,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004245,
      "loss": 6.8413,
      "mean_token_accuracy": 0.09997277483344078,
      "num_tokens": 1480452.0,
      "step": 850
    },
    {
      "entropy": 7.136737203598022,
      "epoch": 0.06652402256370356,
      "grad_norm": 1.078125,
      "learning_rate": 0.000427,
      "loss": 6.7728,
      "mean_token_accuracy": 0.1025382250547409,
      "num_tokens": 1489118.0,
      "step": 855
    },
    {
      "entropy": 7.027253580093384,
      "epoch": 0.06691305193542113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295,
      "loss": 6.7729,
      "mean_token_accuracy": 0.10166823267936706,
      "num_tokens": 1498961.0,
      "step": 860
    },
    {
      "entropy": 7.135442447662354,
      "epoch": 0.06730208130713869,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000432,
      "loss": 6.7701,
      "mean_token_accuracy": 0.10654019340872764,
      "num_tokens": 1506772.0,
      "step": 865
    },
    {
      "entropy": 6.916374588012696,
      "epoch": 0.06769111067885625,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345,
      "loss": 6.6407,
      "mean_token_accuracy": 0.11762609407305717,
      "num_tokens": 1515095.0,
      "step": 870
    },
    {
      "entropy": 6.9736268520355225,
      "epoch": 0.06808014005057382,
      "grad_norm": 1.21875,
      "learning_rate": 0.000437,
      "loss": 6.683,
      "mean_token_accuracy": 0.1091225765645504,
      "num_tokens": 1523477.0,
      "step": 875
    },
    {
      "entropy": 7.0554869174957275,
      "epoch": 0.06846916942229138,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004395,
      "loss": 6.7139,
      "mean_token_accuracy": 0.10232555940747261,
      "num_tokens": 1531786.0,
      "step": 880
    },
    {
      "entropy": 6.96433653831482,
      "epoch": 0.06885819879400895,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000442,
      "loss": 6.8101,
      "mean_token_accuracy": 0.09944222122430801,
      "num_tokens": 1540737.0,
      "step": 885
    },
    {
      "entropy": 6.9922035217285154,
      "epoch": 0.06924722816572651,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004445,
      "loss": 6.6588,
      "mean_token_accuracy": 0.11068925410509109,
      "num_tokens": 1548923.0,
      "step": 890
    },
    {
      "entropy": 6.973821449279785,
      "epoch": 0.06963625753744408,
      "grad_norm": 1.28125,
      "learning_rate": 0.000447,
      "loss": 6.6204,
      "mean_token_accuracy": 0.11367354318499565,
      "num_tokens": 1557043.0,
      "step": 895
    },
    {
      "entropy": 6.884301996231079,
      "epoch": 0.07002528690916164,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.676,
      "mean_token_accuracy": 0.11100396364927292,
      "num_tokens": 1564994.0,
      "step": 900
    },
    {
      "entropy": 6.9740638256073,
      "epoch": 0.0704143162808792,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.6743,
      "mean_token_accuracy": 0.10693940743803979,
      "num_tokens": 1573859.0,
      "step": 905
    },
    {
      "entropy": 6.9993712425231935,
      "epoch": 0.07080334565259677,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.6378,
      "mean_token_accuracy": 0.1043596014380455,
      "num_tokens": 1581947.0,
      "step": 910
    },
    {
      "entropy": 6.931872034072876,
      "epoch": 0.07119237502431433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.6297,
      "mean_token_accuracy": 0.10515054538846016,
      "num_tokens": 1589971.0,
      "step": 915
    },
    {
      "entropy": 6.943262052536011,
      "epoch": 0.0715814043960319,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.5682,
      "mean_token_accuracy": 0.112590941041708,
      "num_tokens": 1598612.0,
      "step": 920
    },
    {
      "entropy": 6.923539781570435,
      "epoch": 0.07197043376774946,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000462,
      "loss": 6.7038,
      "mean_token_accuracy": 0.10295232012867928,
      "num_tokens": 1607378.0,
      "step": 925
    },
    {
      "entropy": 6.933366394042968,
      "epoch": 0.07235946313946703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004645,
      "loss": 6.7097,
      "mean_token_accuracy": 0.10812588483095169,
      "num_tokens": 1615987.0,
      "step": 930
    },
    {
      "entropy": 6.845373868942261,
      "epoch": 0.07274849251118459,
      "grad_norm": 1.140625,
      "learning_rate": 0.000467,
      "loss": 6.5179,
      "mean_token_accuracy": 0.11268422231078148,
      "num_tokens": 1625366.0,
      "step": 935
    },
    {
      "entropy": 6.906345510482788,
      "epoch": 0.07313752188290217,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004695,
      "loss": 6.684,
      "mean_token_accuracy": 0.10814397260546685,
      "num_tokens": 1633020.0,
      "step": 940
    },
    {
      "entropy": 6.9255390644073485,
      "epoch": 0.07352655125461972,
      "grad_norm": 1.0625,
      "learning_rate": 0.000472,
      "loss": 6.6024,
      "mean_token_accuracy": 0.10560433045029641,
      "num_tokens": 1641518.0,
      "step": 945
    },
    {
      "entropy": 6.891413450241089,
      "epoch": 0.07391558062633728,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004745,
      "loss": 6.665,
      "mean_token_accuracy": 0.10579808205366134,
      "num_tokens": 1650472.0,
      "step": 950
    },
    {
      "entropy": 6.896768856048584,
      "epoch": 0.07430460999805485,
      "grad_norm": 1.09375,
      "learning_rate": 0.000477,
      "loss": 6.6856,
      "mean_token_accuracy": 0.10723391622304916,
      "num_tokens": 1659037.0,
      "step": 955
    },
    {
      "entropy": 7.026918506622314,
      "epoch": 0.07469363936977241,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004795,
      "loss": 6.7696,
      "mean_token_accuracy": 0.10622313320636749,
      "num_tokens": 1667897.0,
      "step": 960
    },
    {
      "entropy": 6.8374017715454105,
      "epoch": 0.07508266874148999,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000482,
      "loss": 6.6141,
      "mean_token_accuracy": 0.1083839975297451,
      "num_tokens": 1676535.0,
      "step": 965
    },
    {
      "entropy": 6.900466632843018,
      "epoch": 0.07547169811320754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004845,
      "loss": 6.6481,
      "mean_token_accuracy": 0.10983363538980484,
      "num_tokens": 1684415.0,
      "step": 970
    },
    {
      "entropy": 6.9876219749450685,
      "epoch": 0.07586072748492512,
      "grad_norm": 1.078125,
      "learning_rate": 0.000487,
      "loss": 6.6023,
      "mean_token_accuracy": 0.11541447341442108,
      "num_tokens": 1692443.0,
      "step": 975
    },
    {
      "entropy": 6.73767499923706,
      "epoch": 0.07624975685664268,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895,
      "loss": 6.6032,
      "mean_token_accuracy": 0.1102523811161518,
      "num_tokens": 1700814.0,
      "step": 980
    },
    {
      "entropy": 6.8459278583526615,
      "epoch": 0.07663878622836025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492,
      "loss": 6.4589,
      "mean_token_accuracy": 0.11454312652349471,
      "num_tokens": 1709591.0,
      "step": 985
    },
    {
      "entropy": 6.855618333816528,
      "epoch": 0.0770278156000778,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945,
      "loss": 6.6899,
      "mean_token_accuracy": 0.10243948996067047,
      "num_tokens": 1718951.0,
      "step": 990
    },
    {
      "entropy": 7.0016392230987545,
      "epoch": 0.07741684497179536,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000497,
      "loss": 6.6382,
      "mean_token_accuracy": 0.10815301239490509,
      "num_tokens": 1727481.0,
      "step": 995
    },
    {
      "entropy": 6.801933670043946,
      "epoch": 0.07780587434351294,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995,
      "loss": 6.6336,
      "mean_token_accuracy": 0.10878749638795852,
      "num_tokens": 1736985.0,
      "step": 1000
    },
    {
      "entropy": 6.949926328659058,
      "epoch": 0.0781949037152305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999989075146,
      "loss": 6.6868,
      "mean_token_accuracy": 0.10359607338905334,
      "num_tokens": 1745470.0,
      "step": 1005
    },
    {
      "entropy": 6.897040796279907,
      "epoch": 0.07858393308694807,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999999944692927,
      "loss": 6.629,
      "mean_token_accuracy": 0.10731491297483445,
      "num_tokens": 1754185.0,
      "step": 1010
    },
    {
      "entropy": 6.833609628677368,
      "epoch": 0.07897296245866563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499999986617054,
      "loss": 6.6342,
      "mean_token_accuracy": 0.10663639605045319,
      "num_tokens": 1762100.0,
      "step": 1015
    },
    {
      "entropy": 6.971498203277588,
      "epoch": 0.0793619918303832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999753507986,
      "loss": 6.6379,
      "mean_token_accuracy": 0.10761738270521164,
      "num_tokens": 1771070.0,
      "step": 1020
    },
    {
      "entropy": 6.819230461120606,
      "epoch": 0.07975102120210076,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999999606705267,
      "loss": 6.6338,
      "mean_token_accuracy": 0.10975765809416771,
      "num_tokens": 1780085.0,
      "step": 1025
    },
    {
      "entropy": 6.906311321258545,
      "epoch": 0.08014005057381832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999425762385,
      "loss": 6.6172,
      "mean_token_accuracy": 0.113339963555336,
      "num_tokens": 1789316.0,
      "step": 1030
    },
    {
      "entropy": 6.783714866638183,
      "epoch": 0.08052907994553589,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999210679344,
      "loss": 6.5369,
      "mean_token_accuracy": 0.112546556442976,
      "num_tokens": 1797975.0,
      "step": 1035
    },
    {
      "entropy": 6.838487339019776,
      "epoch": 0.08091810931725345,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999998961456145,
      "loss": 6.6393,
      "mean_token_accuracy": 0.11156408786773682,
      "num_tokens": 1806129.0,
      "step": 1040
    },
    {
      "entropy": 6.908475303649903,
      "epoch": 0.08130713868897102,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999998678092794,
      "loss": 6.6392,
      "mean_token_accuracy": 0.11067313626408577,
      "num_tokens": 1813956.0,
      "step": 1045
    },
    {
      "entropy": 6.9355841159820555,
      "epoch": 0.08169616806068858,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999998360589293,
      "loss": 6.6473,
      "mean_token_accuracy": 0.11026927605271339,
      "num_tokens": 1822572.0,
      "step": 1050
    },
    {
      "entropy": 6.757045555114746,
      "epoch": 0.08208519743240615,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499999800894565,
      "loss": 6.5619,
      "mean_token_accuracy": 0.11012868583202362,
      "num_tokens": 1831870.0,
      "step": 1055
    },
    {
      "entropy": 6.749943971633911,
      "epoch": 0.08247422680412371,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999997623161866,
      "loss": 6.593,
      "mean_token_accuracy": 0.10753906518220901,
      "num_tokens": 1840414.0,
      "step": 1060
    },
    {
      "entropy": 6.898047304153442,
      "epoch": 0.08286325617584128,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999997203237951,
      "loss": 6.6142,
      "mean_token_accuracy": 0.10993773117661476,
      "num_tokens": 1850385.0,
      "step": 1065
    },
    {
      "entropy": 6.692205476760864,
      "epoch": 0.08325228554755884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499999674917391,
      "loss": 6.5192,
      "mean_token_accuracy": 0.12006976082921028,
      "num_tokens": 1858479.0,
      "step": 1070
    },
    {
      "entropy": 6.8471941471099855,
      "epoch": 0.0836413149192764,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999996260969748,
      "loss": 6.5988,
      "mean_token_accuracy": 0.10776122882962227,
      "num_tokens": 1867357.0,
      "step": 1075
    },
    {
      "entropy": 6.78913459777832,
      "epoch": 0.08403034429099397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999995738625475,
      "loss": 6.5436,
      "mean_token_accuracy": 0.10430336892604827,
      "num_tokens": 1876211.0,
      "step": 1080
    },
    {
      "entropy": 6.690899896621704,
      "epoch": 0.08441937366271153,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999995182141099,
      "loss": 6.4663,
      "mean_token_accuracy": 0.10824713706970215,
      "num_tokens": 1884301.0,
      "step": 1085
    },
    {
      "entropy": 6.911795234680175,
      "epoch": 0.0848084030344291,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999994591516626,
      "loss": 6.622,
      "mean_token_accuracy": 0.11672904044389724,
      "num_tokens": 1893191.0,
      "step": 1090
    },
    {
      "entropy": 6.69744906425476,
      "epoch": 0.08519743240614666,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999993966752066,
      "loss": 6.4871,
      "mean_token_accuracy": 0.11046854481101036,
      "num_tokens": 1901434.0,
      "step": 1095
    },
    {
      "entropy": 6.7617669105529785,
      "epoch": 0.08558646177786423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499999330784743,
      "loss": 6.4873,
      "mean_token_accuracy": 0.1202208161354065,
      "num_tokens": 1910054.0,
      "step": 1100
    },
    {
      "entropy": 6.82279953956604,
      "epoch": 0.08597549114958179,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999992614802725,
      "loss": 6.6178,
      "mean_token_accuracy": 0.11331576108932495,
      "num_tokens": 1918607.0,
      "step": 1105
    },
    {
      "entropy": 6.74808177947998,
      "epoch": 0.08636452052129936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999991887617966,
      "loss": 6.5604,
      "mean_token_accuracy": 0.11329781040549278,
      "num_tokens": 1927610.0,
      "step": 1110
    },
    {
      "entropy": 6.721193456649781,
      "epoch": 0.08675354989301692,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999991126293158,
      "loss": 6.5076,
      "mean_token_accuracy": 0.11460185199975967,
      "num_tokens": 1936174.0,
      "step": 1115
    },
    {
      "entropy": 6.726939487457275,
      "epoch": 0.08714257926473448,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999990330828318,
      "loss": 6.437,
      "mean_token_accuracy": 0.12100310698151588,
      "num_tokens": 1944228.0,
      "step": 1120
    },
    {
      "entropy": 6.830557537078858,
      "epoch": 0.08753160863645205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999989501223456,
      "loss": 6.5437,
      "mean_token_accuracy": 0.11512776464223862,
      "num_tokens": 1953144.0,
      "step": 1125
    },
    {
      "entropy": 6.717128944396973,
      "epoch": 0.08792063800816961,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999988637478584,
      "loss": 6.5947,
      "mean_token_accuracy": 0.11105897948145867,
      "num_tokens": 1961947.0,
      "step": 1130
    },
    {
      "entropy": 6.843059253692627,
      "epoch": 0.08830966737988719,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999987739593716,
      "loss": 6.5708,
      "mean_token_accuracy": 0.11229311525821686,
      "num_tokens": 1970674.0,
      "step": 1135
    },
    {
      "entropy": 6.781560564041138,
      "epoch": 0.08869869675160474,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999986807568865,
      "loss": 6.5356,
      "mean_token_accuracy": 0.11566499695181846,
      "num_tokens": 1979173.0,
      "step": 1140
    },
    {
      "entropy": 6.708818435668945,
      "epoch": 0.08908772612332232,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999985841404045,
      "loss": 6.5334,
      "mean_token_accuracy": 0.11458146646618843,
      "num_tokens": 1988134.0,
      "step": 1145
    },
    {
      "entropy": 6.784994983673096,
      "epoch": 0.08947675549503988,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999984841099271,
      "loss": 6.4755,
      "mean_token_accuracy": 0.11356236785650253,
      "num_tokens": 1996977.0,
      "step": 1150
    },
    {
      "entropy": 6.741352796554565,
      "epoch": 0.08986578486675743,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499998380665456,
      "loss": 6.5366,
      "mean_token_accuracy": 0.11298326328396797,
      "num_tokens": 2005946.0,
      "step": 1155
    },
    {
      "entropy": 6.695943164825439,
      "epoch": 0.090254814238475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999982738069924,
      "loss": 6.3749,
      "mean_token_accuracy": 0.11534976065158845,
      "num_tokens": 2014882.0,
      "step": 1160
    },
    {
      "entropy": 6.682829093933106,
      "epoch": 0.09064384361019256,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999981635345382,
      "loss": 6.5476,
      "mean_token_accuracy": 0.11290719136595725,
      "num_tokens": 2024858.0,
      "step": 1165
    },
    {
      "entropy": 6.672901630401611,
      "epoch": 0.09103287298191014,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999980498480949,
      "loss": 6.466,
      "mean_token_accuracy": 0.11438599750399589,
      "num_tokens": 2034072.0,
      "step": 1170
    },
    {
      "entropy": 6.712941026687622,
      "epoch": 0.0914219023536277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999979327476644,
      "loss": 6.4753,
      "mean_token_accuracy": 0.11522193774580955,
      "num_tokens": 2042649.0,
      "step": 1175
    },
    {
      "entropy": 6.714944648742676,
      "epoch": 0.09181093172534527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999978122332485,
      "loss": 6.4979,
      "mean_token_accuracy": 0.11657196953892708,
      "num_tokens": 2050866.0,
      "step": 1180
    },
    {
      "entropy": 6.751889419555664,
      "epoch": 0.09219996109706283,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999976883048487,
      "loss": 6.4984,
      "mean_token_accuracy": 0.1147743172943592,
      "num_tokens": 2058914.0,
      "step": 1185
    },
    {
      "entropy": 6.674278974533081,
      "epoch": 0.0925889904687804,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999975609624673,
      "loss": 6.4469,
      "mean_token_accuracy": 0.11601505503058433,
      "num_tokens": 2068098.0,
      "step": 1190
    },
    {
      "entropy": 6.666662788391113,
      "epoch": 0.09297801984049796,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999974302061059,
      "loss": 6.4465,
      "mean_token_accuracy": 0.11012001484632492,
      "num_tokens": 2077299.0,
      "step": 1195
    },
    {
      "entropy": 6.718023204803467,
      "epoch": 0.09336704921221552,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999972960357666,
      "loss": 6.5418,
      "mean_token_accuracy": 0.11328572109341621,
      "num_tokens": 2087868.0,
      "step": 1200
    },
    {
      "entropy": 6.68776216506958,
      "epoch": 0.09375607858393309,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999971584514516,
      "loss": 6.4704,
      "mean_token_accuracy": 0.12280796691775323,
      "num_tokens": 2096553.0,
      "step": 1205
    },
    {
      "entropy": 6.763374853134155,
      "epoch": 0.09414510795565065,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999970174531627,
      "loss": 6.4254,
      "mean_token_accuracy": 0.11939279362559319,
      "num_tokens": 2104809.0,
      "step": 1210
    },
    {
      "entropy": 6.544530963897705,
      "epoch": 0.09453413732736822,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999968730409022,
      "loss": 6.4114,
      "mean_token_accuracy": 0.11297131180763245,
      "num_tokens": 2113869.0,
      "step": 1215
    },
    {
      "entropy": 6.637964534759521,
      "epoch": 0.09492316669908578,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999967252146723,
      "loss": 6.245,
      "mean_token_accuracy": 0.1171722136437893,
      "num_tokens": 2121749.0,
      "step": 1220
    },
    {
      "entropy": 6.630059051513672,
      "epoch": 0.09531219607080335,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999965739744752,
      "loss": 6.4551,
      "mean_token_accuracy": 0.11625204831361771,
      "num_tokens": 2130828.0,
      "step": 1225
    },
    {
      "entropy": 6.729389619827271,
      "epoch": 0.09570122544252091,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999964193203132,
      "loss": 6.4343,
      "mean_token_accuracy": 0.11840956956148148,
      "num_tokens": 2138928.0,
      "step": 1230
    },
    {
      "entropy": 6.677809381484986,
      "epoch": 0.09609025481423847,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999962612521888,
      "loss": 6.4553,
      "mean_token_accuracy": 0.11565364599227905,
      "num_tokens": 2147643.0,
      "step": 1235
    },
    {
      "entropy": 6.696507167816162,
      "epoch": 0.09647928418595604,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999960997701041,
      "loss": 6.3968,
      "mean_token_accuracy": 0.1201579064130783,
      "num_tokens": 2156766.0,
      "step": 1240
    },
    {
      "entropy": 6.621823453903199,
      "epoch": 0.0968683135576736,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999959348740618,
      "loss": 6.4417,
      "mean_token_accuracy": 0.11661816611886025,
      "num_tokens": 2165517.0,
      "step": 1245
    },
    {
      "entropy": 6.576675033569336,
      "epoch": 0.09725734292939117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999957665640641,
      "loss": 6.3206,
      "mean_token_accuracy": 0.12200571298599243,
      "num_tokens": 2174812.0,
      "step": 1250
    },
    {
      "entropy": 6.646194124221802,
      "epoch": 0.09764637230110873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499995594840114,
      "loss": 6.4638,
      "mean_token_accuracy": 0.12040529623627663,
      "num_tokens": 2183968.0,
      "step": 1255
    },
    {
      "entropy": 6.681028127670288,
      "epoch": 0.0980354016728263,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999954197022138,
      "loss": 6.4909,
      "mean_token_accuracy": 0.1184639185667038,
      "num_tokens": 2191947.0,
      "step": 1260
    },
    {
      "entropy": 6.642876625061035,
      "epoch": 0.09842443104454386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999952411503661,
      "loss": 6.3293,
      "mean_token_accuracy": 0.12336919456720352,
      "num_tokens": 2199951.0,
      "step": 1265
    },
    {
      "entropy": 6.591861581802368,
      "epoch": 0.09881346041626143,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999950591845738,
      "loss": 6.4673,
      "mean_token_accuracy": 0.11935944855213165,
      "num_tokens": 2208714.0,
      "step": 1270
    },
    {
      "entropy": 6.581157541275024,
      "epoch": 0.09920248978797899,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999948738048397,
      "loss": 6.3282,
      "mean_token_accuracy": 0.11878729537129402,
      "num_tokens": 2217250.0,
      "step": 1275
    },
    {
      "entropy": 6.7235880374908445,
      "epoch": 0.09959151915969655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999946850111663,
      "loss": 6.3894,
      "mean_token_accuracy": 0.12055943980813026,
      "num_tokens": 2225468.0,
      "step": 1280
    },
    {
      "entropy": 6.508683109283448,
      "epoch": 0.09998054853141412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999944928035569,
      "loss": 6.3437,
      "mean_token_accuracy": 0.12606204822659492,
      "num_tokens": 2233893.0,
      "step": 1285
    },
    {
      "entropy": 6.760361814498902,
      "epoch": 0.10036957790313168,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499994297182014,
      "loss": 6.4239,
      "mean_token_accuracy": 0.11900506541132927,
      "num_tokens": 2242487.0,
      "step": 1290
    },
    {
      "entropy": 6.464075803756714,
      "epoch": 0.10075860727484925,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999940981465408,
      "loss": 6.2418,
      "mean_token_accuracy": 0.12773674577474595,
      "num_tokens": 2250518.0,
      "step": 1295
    },
    {
      "entropy": 6.635386562347412,
      "epoch": 0.10114763664656681,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999938956971404,
      "loss": 6.4266,
      "mean_token_accuracy": 0.11632013618946076,
      "num_tokens": 2259024.0,
      "step": 1300
    },
    {
      "entropy": 6.615658807754516,
      "epoch": 0.10153666601828439,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999936898338156,
      "loss": 6.4632,
      "mean_token_accuracy": 0.11199147924780846,
      "num_tokens": 2267832.0,
      "step": 1305
    },
    {
      "entropy": 6.635924625396728,
      "epoch": 0.10192569539000194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999934805565698,
      "loss": 6.5261,
      "mean_token_accuracy": 0.11091560125350952,
      "num_tokens": 2276834.0,
      "step": 1310
    },
    {
      "entropy": 6.73242449760437,
      "epoch": 0.10231472476171952,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499993267865406,
      "loss": 6.4461,
      "mean_token_accuracy": 0.11519345194101334,
      "num_tokens": 2285736.0,
      "step": 1315
    },
    {
      "entropy": 6.697244501113891,
      "epoch": 0.10270375413343708,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999930517603275,
      "loss": 6.4108,
      "mean_token_accuracy": 0.1263464353978634,
      "num_tokens": 2294177.0,
      "step": 1320
    },
    {
      "entropy": 6.468812894821167,
      "epoch": 0.10309278350515463,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999928322413374,
      "loss": 6.1839,
      "mean_token_accuracy": 0.12833936586976052,
      "num_tokens": 2302784.0,
      "step": 1325
    },
    {
      "entropy": 6.505994129180908,
      "epoch": 0.1034818128768722,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999926093084394,
      "loss": 6.2968,
      "mean_token_accuracy": 0.12245660051703453,
      "num_tokens": 2311650.0,
      "step": 1330
    },
    {
      "entropy": 6.66971116065979,
      "epoch": 0.10387084224858976,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999923829616365,
      "loss": 6.3996,
      "mean_token_accuracy": 0.11883401051163674,
      "num_tokens": 2319828.0,
      "step": 1335
    },
    {
      "entropy": 6.553110504150391,
      "epoch": 0.10425987162030734,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999921532009323,
      "loss": 6.416,
      "mean_token_accuracy": 0.11476577743887902,
      "num_tokens": 2328636.0,
      "step": 1340
    },
    {
      "entropy": 6.5455653190612795,
      "epoch": 0.1046489009920249,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999919200263304,
      "loss": 6.2977,
      "mean_token_accuracy": 0.12469864264130592,
      "num_tokens": 2336771.0,
      "step": 1345
    },
    {
      "entropy": 6.624483728408814,
      "epoch": 0.10503793036374247,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999916834378342,
      "loss": 6.4529,
      "mean_token_accuracy": 0.11439902409911155,
      "num_tokens": 2346106.0,
      "step": 1350
    },
    {
      "entropy": 6.588077735900879,
      "epoch": 0.10542695973546003,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999914434354472,
      "loss": 6.4435,
      "mean_token_accuracy": 0.11157507598400115,
      "num_tokens": 2354943.0,
      "step": 1355
    },
    {
      "entropy": 6.484687662124633,
      "epoch": 0.10581598910717759,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999912000191733,
      "loss": 6.3291,
      "mean_token_accuracy": 0.12155802845954895,
      "num_tokens": 2363699.0,
      "step": 1360
    },
    {
      "entropy": 6.56774754524231,
      "epoch": 0.10620501847889516,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499990953189016,
      "loss": 6.4607,
      "mean_token_accuracy": 0.11734867691993714,
      "num_tokens": 2373316.0,
      "step": 1365
    },
    {
      "entropy": 6.792719316482544,
      "epoch": 0.10659404785061272,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999907029449791,
      "loss": 6.5466,
      "mean_token_accuracy": 0.1130414329469204,
      "num_tokens": 2381550.0,
      "step": 1370
    },
    {
      "entropy": 6.568495798110962,
      "epoch": 0.10698307722233029,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999904492870663,
      "loss": 6.3407,
      "mean_token_accuracy": 0.12280402705073357,
      "num_tokens": 2389546.0,
      "step": 1375
    },
    {
      "entropy": 6.602413702011108,
      "epoch": 0.10737210659404785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999901922152816,
      "loss": 6.3119,
      "mean_token_accuracy": 0.12875628471374512,
      "num_tokens": 2398852.0,
      "step": 1380
    },
    {
      "entropy": 6.514067602157593,
      "epoch": 0.10776113596576542,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999899317296288,
      "loss": 6.4055,
      "mean_token_accuracy": 0.11861231550574303,
      "num_tokens": 2407815.0,
      "step": 1385
    },
    {
      "entropy": 6.577896213531494,
      "epoch": 0.10815016533748298,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999896678301121,
      "loss": 6.2648,
      "mean_token_accuracy": 0.12172571048140526,
      "num_tokens": 2417300.0,
      "step": 1390
    },
    {
      "entropy": 6.5998763084411625,
      "epoch": 0.10853919470920055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999894005167352,
      "loss": 6.4098,
      "mean_token_accuracy": 0.12156257852911949,
      "num_tokens": 2426257.0,
      "step": 1395
    },
    {
      "entropy": 6.676355028152466,
      "epoch": 0.10892822408091811,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499989129789502,
      "loss": 6.4257,
      "mean_token_accuracy": 0.1204957626760006,
      "num_tokens": 2434595.0,
      "step": 1400
    },
    {
      "entropy": 6.654974126815796,
      "epoch": 0.10931725345263567,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999888556484172,
      "loss": 6.4462,
      "mean_token_accuracy": 0.11929797455668449,
      "num_tokens": 2443945.0,
      "step": 1405
    },
    {
      "entropy": 6.505768489837647,
      "epoch": 0.10970628282435324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999885780934845,
      "loss": 6.3098,
      "mean_token_accuracy": 0.12298460751771927,
      "num_tokens": 2453141.0,
      "step": 1410
    },
    {
      "entropy": 6.586579370498657,
      "epoch": 0.1100953121960708,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999882971247081,
      "loss": 6.3959,
      "mean_token_accuracy": 0.11581015586853027,
      "num_tokens": 2461873.0,
      "step": 1415
    },
    {
      "entropy": 6.595294094085693,
      "epoch": 0.11048434156778837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999880127420926,
      "loss": 6.4179,
      "mean_token_accuracy": 0.11764904409646988,
      "num_tokens": 2471238.0,
      "step": 1420
    },
    {
      "entropy": 6.5501360416412355,
      "epoch": 0.11087337093950593,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999877249456421,
      "loss": 6.2916,
      "mean_token_accuracy": 0.1313754253089428,
      "num_tokens": 2479412.0,
      "step": 1425
    },
    {
      "entropy": 6.429438781738281,
      "epoch": 0.1112624003112235,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999874337353609,
      "loss": 6.2078,
      "mean_token_accuracy": 0.1259915679693222,
      "num_tokens": 2488150.0,
      "step": 1430
    },
    {
      "entropy": 6.529017305374145,
      "epoch": 0.11165142968294106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999871391112535,
      "loss": 6.3654,
      "mean_token_accuracy": 0.1272561579942703,
      "num_tokens": 2498103.0,
      "step": 1435
    },
    {
      "entropy": 6.49380464553833,
      "epoch": 0.11204045905465862,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999868410733244,
      "loss": 6.2319,
      "mean_token_accuracy": 0.1280946686863899,
      "num_tokens": 2506211.0,
      "step": 1440
    },
    {
      "entropy": 6.4133141994476315,
      "epoch": 0.11242948842637619,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499986539621578,
      "loss": 6.2552,
      "mean_token_accuracy": 0.1278551071882248,
      "num_tokens": 2514743.0,
      "step": 1445
    },
    {
      "entropy": 6.430079650878906,
      "epoch": 0.11281851779809375,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999862347560191,
      "loss": 6.2533,
      "mean_token_accuracy": 0.12282558158040047,
      "num_tokens": 2523824.0,
      "step": 1450
    },
    {
      "entropy": 6.598657894134521,
      "epoch": 0.11320754716981132,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999859264766521,
      "loss": 6.3219,
      "mean_token_accuracy": 0.12097884491086006,
      "num_tokens": 2532360.0,
      "step": 1455
    },
    {
      "entropy": 6.540658664703369,
      "epoch": 0.11359657654152888,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999856147834817,
      "loss": 6.2807,
      "mean_token_accuracy": 0.13108301162719727,
      "num_tokens": 2540547.0,
      "step": 1460
    },
    {
      "entropy": 6.4696839332580565,
      "epoch": 0.11398560591324645,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999852996765129,
      "loss": 6.3052,
      "mean_token_accuracy": 0.12217096462845803,
      "num_tokens": 2549193.0,
      "step": 1465
    },
    {
      "entropy": 6.500145196914673,
      "epoch": 0.11437463528496401,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999849811557502,
      "loss": 6.3173,
      "mean_token_accuracy": 0.11922707110643387,
      "num_tokens": 2557553.0,
      "step": 1470
    },
    {
      "entropy": 6.638996171951294,
      "epoch": 0.11476366465668159,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999846592211984,
      "loss": 6.4043,
      "mean_token_accuracy": 0.1211747631430626,
      "num_tokens": 2565945.0,
      "step": 1475
    },
    {
      "entropy": 6.573302364349365,
      "epoch": 0.11515269402839914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999843338728626,
      "loss": 6.2786,
      "mean_token_accuracy": 0.11536813229322433,
      "num_tokens": 2574364.0,
      "step": 1480
    },
    {
      "entropy": 6.438839960098266,
      "epoch": 0.1155417234001167,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999840051107476,
      "loss": 6.2133,
      "mean_token_accuracy": 0.12060227617621422,
      "num_tokens": 2582825.0,
      "step": 1485
    },
    {
      "entropy": 6.600337505340576,
      "epoch": 0.11593075277183428,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999836729348584,
      "loss": 6.3589,
      "mean_token_accuracy": 0.12052287831902504,
      "num_tokens": 2591628.0,
      "step": 1490
    },
    {
      "entropy": 6.495076751708984,
      "epoch": 0.11631978214355183,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999833373452,
      "loss": 6.3561,
      "mean_token_accuracy": 0.1171538695693016,
      "num_tokens": 2600436.0,
      "step": 1495
    },
    {
      "entropy": 6.521013927459717,
      "epoch": 0.1167088115152694,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999829983417778,
      "loss": 6.299,
      "mean_token_accuracy": 0.12321334183216096,
      "num_tokens": 2609709.0,
      "step": 1500
    },
    {
      "entropy": 6.610712099075317,
      "epoch": 0.11709784088698696,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999826559245965,
      "loss": 6.3006,
      "mean_token_accuracy": 0.12215944081544876,
      "num_tokens": 2617992.0,
      "step": 1505
    },
    {
      "entropy": 6.396210336685181,
      "epoch": 0.11748687025870454,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999823100936615,
      "loss": 6.2158,
      "mean_token_accuracy": 0.12474563270807267,
      "num_tokens": 2626706.0,
      "step": 1510
    },
    {
      "entropy": 6.615616989135742,
      "epoch": 0.1178758996304221,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999819608489781,
      "loss": 6.287,
      "mean_token_accuracy": 0.12880871593952178,
      "num_tokens": 2634642.0,
      "step": 1515
    },
    {
      "entropy": 6.544697713851929,
      "epoch": 0.11826492900213967,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999816081905515,
      "loss": 6.3346,
      "mean_token_accuracy": 0.11531090214848519,
      "num_tokens": 2644160.0,
      "step": 1520
    },
    {
      "entropy": 6.434164953231812,
      "epoch": 0.11865395837385723,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999812521183872,
      "loss": 6.2948,
      "mean_token_accuracy": 0.11871815845370293,
      "num_tokens": 2653390.0,
      "step": 1525
    },
    {
      "entropy": 6.589237403869629,
      "epoch": 0.11904298774557479,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999808926324903,
      "loss": 6.3435,
      "mean_token_accuracy": 0.12100820019841194,
      "num_tokens": 2662561.0,
      "step": 1530
    },
    {
      "entropy": 6.478472805023193,
      "epoch": 0.11943201711729236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999805297328665,
      "loss": 6.3086,
      "mean_token_accuracy": 0.11910761371254922,
      "num_tokens": 2670694.0,
      "step": 1535
    },
    {
      "entropy": 6.46109447479248,
      "epoch": 0.11982104648900992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999801634195214,
      "loss": 6.1965,
      "mean_token_accuracy": 0.12499032095074654,
      "num_tokens": 2679583.0,
      "step": 1540
    },
    {
      "entropy": 6.4330726146698,
      "epoch": 0.12021007586072749,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999797936924603,
      "loss": 6.1847,
      "mean_token_accuracy": 0.13039473071694374,
      "num_tokens": 2688429.0,
      "step": 1545
    },
    {
      "entropy": 6.5682470321655275,
      "epoch": 0.12059910523244505,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499979420551689,
      "loss": 6.258,
      "mean_token_accuracy": 0.12565399780869485,
      "num_tokens": 2697134.0,
      "step": 1550
    },
    {
      "entropy": 6.446918916702271,
      "epoch": 0.12098813460416262,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499979043997213,
      "loss": 6.4093,
      "mean_token_accuracy": 0.11900994554162025,
      "num_tokens": 2705641.0,
      "step": 1555
    },
    {
      "entropy": 6.5380446434021,
      "epoch": 0.12137716397588018,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999786640290381,
      "loss": 6.333,
      "mean_token_accuracy": 0.1262545704841614,
      "num_tokens": 2714977.0,
      "step": 1560
    },
    {
      "entropy": 6.517403984069825,
      "epoch": 0.12176619334759774,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999782806471701,
      "loss": 6.1801,
      "mean_token_accuracy": 0.13938897997140884,
      "num_tokens": 2723120.0,
      "step": 1565
    },
    {
      "entropy": 6.311800098419189,
      "epoch": 0.12215522271931531,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999778938516147,
      "loss": 6.1791,
      "mean_token_accuracy": 0.1359429642558098,
      "num_tokens": 2731964.0,
      "step": 1570
    },
    {
      "entropy": 6.646980047225952,
      "epoch": 0.12254425209103287,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499977503642378,
      "loss": 6.2559,
      "mean_token_accuracy": 0.1213235966861248,
      "num_tokens": 2741537.0,
      "step": 1575
    },
    {
      "entropy": 6.3805046558380125,
      "epoch": 0.12293328146275044,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999771100194656,
      "loss": 6.3638,
      "mean_token_accuracy": 0.12285436391830444,
      "num_tokens": 2751522.0,
      "step": 1580
    },
    {
      "entropy": 6.59956202507019,
      "epoch": 0.123322310834468,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999767129828837,
      "loss": 6.3669,
      "mean_token_accuracy": 0.11640789434313774,
      "num_tokens": 2761100.0,
      "step": 1585
    },
    {
      "entropy": 6.441359281539917,
      "epoch": 0.12371134020618557,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999763125326383,
      "loss": 6.2932,
      "mean_token_accuracy": 0.1239714004099369,
      "num_tokens": 2769171.0,
      "step": 1590
    },
    {
      "entropy": 6.480400943756104,
      "epoch": 0.12410036957790313,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999759086687355,
      "loss": 6.2306,
      "mean_token_accuracy": 0.12737515345215797,
      "num_tokens": 2778021.0,
      "step": 1595
    },
    {
      "entropy": 6.3953241348266605,
      "epoch": 0.1244893989496207,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999755013911813,
      "loss": 6.1706,
      "mean_token_accuracy": 0.12780368328094482,
      "num_tokens": 2787206.0,
      "step": 1600
    },
    {
      "entropy": 6.448420667648316,
      "epoch": 0.12487842832133826,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499975090699982,
      "loss": 6.2372,
      "mean_token_accuracy": 0.12469636872410775,
      "num_tokens": 2796051.0,
      "step": 1605
    },
    {
      "entropy": 6.501862096786499,
      "epoch": 0.12526745769305583,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999746765951438,
      "loss": 6.1934,
      "mean_token_accuracy": 0.12748273462057114,
      "num_tokens": 2804961.0,
      "step": 1610
    },
    {
      "entropy": 6.395801019668579,
      "epoch": 0.1256564870647734,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999742590766729,
      "loss": 6.1504,
      "mean_token_accuracy": 0.12995274290442466,
      "num_tokens": 2813691.0,
      "step": 1615
    },
    {
      "entropy": 6.438721466064453,
      "epoch": 0.12604551643649095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999738381445757,
      "loss": 6.2961,
      "mean_token_accuracy": 0.12733198031783105,
      "num_tokens": 2822248.0,
      "step": 1620
    },
    {
      "entropy": 6.4921409606933596,
      "epoch": 0.1264345458082085,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999734137988586,
      "loss": 6.2473,
      "mean_token_accuracy": 0.12927133068442345,
      "num_tokens": 2830826.0,
      "step": 1625
    },
    {
      "entropy": 6.472244501113892,
      "epoch": 0.1268235751799261,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499972986039528,
      "loss": 6.2285,
      "mean_token_accuracy": 0.1285054236650467,
      "num_tokens": 2839766.0,
      "step": 1630
    },
    {
      "entropy": 6.295017290115356,
      "epoch": 0.12721260455164365,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999725548665904,
      "loss": 6.1702,
      "mean_token_accuracy": 0.13327566310763359,
      "num_tokens": 2848728.0,
      "step": 1635
    },
    {
      "entropy": 6.369825220108032,
      "epoch": 0.1276016339233612,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999721202800524,
      "loss": 6.1868,
      "mean_token_accuracy": 0.13074957728385925,
      "num_tokens": 2857221.0,
      "step": 1640
    },
    {
      "entropy": 6.546764755249024,
      "epoch": 0.12799066329507877,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999716822799206,
      "loss": 6.2481,
      "mean_token_accuracy": 0.13030916303396226,
      "num_tokens": 2865314.0,
      "step": 1645
    },
    {
      "entropy": 6.320151281356812,
      "epoch": 0.12837969266679633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999712408662016,
      "loss": 6.1424,
      "mean_token_accuracy": 0.12682838290929793,
      "num_tokens": 2873654.0,
      "step": 1650
    },
    {
      "entropy": 6.379095506668091,
      "epoch": 0.12876872203851392,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999707960389021,
      "loss": 6.0377,
      "mean_token_accuracy": 0.13998508602380752,
      "num_tokens": 2881604.0,
      "step": 1655
    },
    {
      "entropy": 6.374504232406617,
      "epoch": 0.12915775141023147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999703477980288,
      "loss": 6.2388,
      "mean_token_accuracy": 0.12339747697114944,
      "num_tokens": 2890427.0,
      "step": 1660
    },
    {
      "entropy": 6.471899557113647,
      "epoch": 0.12954678078194903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999698961435885,
      "loss": 6.2024,
      "mean_token_accuracy": 0.12862686440348625,
      "num_tokens": 2899206.0,
      "step": 1665
    },
    {
      "entropy": 6.3578918933868405,
      "epoch": 0.1299358101536666,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999694410755882,
      "loss": 6.1251,
      "mean_token_accuracy": 0.13412127643823624,
      "num_tokens": 2907738.0,
      "step": 1670
    },
    {
      "entropy": 6.475336980819702,
      "epoch": 0.13032483952538418,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999689825940347,
      "loss": 6.3526,
      "mean_token_accuracy": 0.12458359599113464,
      "num_tokens": 2916881.0,
      "step": 1675
    },
    {
      "entropy": 6.48715295791626,
      "epoch": 0.13071386889710174,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999685206989349,
      "loss": 6.2833,
      "mean_token_accuracy": 0.12076244428753853,
      "num_tokens": 2926467.0,
      "step": 1680
    },
    {
      "entropy": 6.485057735443116,
      "epoch": 0.1311028982688193,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999680553902959,
      "loss": 6.2095,
      "mean_token_accuracy": 0.12537256702780725,
      "num_tokens": 2934671.0,
      "step": 1685
    },
    {
      "entropy": 6.374607038497925,
      "epoch": 0.13149192764053685,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999675866681247,
      "loss": 6.2382,
      "mean_token_accuracy": 0.12314836755394935,
      "num_tokens": 2943925.0,
      "step": 1690
    },
    {
      "entropy": 6.529426240921021,
      "epoch": 0.1318809570122544,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999671145324286,
      "loss": 6.2839,
      "mean_token_accuracy": 0.12293665111064911,
      "num_tokens": 2952118.0,
      "step": 1695
    },
    {
      "entropy": 6.332964897155762,
      "epoch": 0.132269986383972,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999666389832144,
      "loss": 6.246,
      "mean_token_accuracy": 0.12483881637454033,
      "num_tokens": 2961110.0,
      "step": 1700
    },
    {
      "entropy": 6.501490259170533,
      "epoch": 0.13265901575568956,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999661600204897,
      "loss": 6.1969,
      "mean_token_accuracy": 0.1331293873488903,
      "num_tokens": 2968877.0,
      "step": 1705
    },
    {
      "entropy": 6.432631254196167,
      "epoch": 0.13304804512740712,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999656776442615,
      "loss": 6.2441,
      "mean_token_accuracy": 0.12796934992074965,
      "num_tokens": 2977346.0,
      "step": 1710
    },
    {
      "entropy": 6.520103311538696,
      "epoch": 0.13343707449912467,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999651918545372,
      "loss": 6.2982,
      "mean_token_accuracy": 0.12270353958010674,
      "num_tokens": 2987494.0,
      "step": 1715
    },
    {
      "entropy": 6.414762353897094,
      "epoch": 0.13382610387084226,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999647026513243,
      "loss": 6.2018,
      "mean_token_accuracy": 0.12620207220315932,
      "num_tokens": 2996728.0,
      "step": 1720
    },
    {
      "entropy": 6.426468324661255,
      "epoch": 0.13421513324255982,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049996421003463,
      "loss": 6.2245,
      "mean_token_accuracy": 0.12518492266535758,
      "num_tokens": 3005810.0,
      "step": 1725
    },
    {
      "entropy": 6.418255186080932,
      "epoch": 0.13460416261427738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999637140044619,
      "loss": 6.2506,
      "mean_token_accuracy": 0.12523002028465272,
      "num_tokens": 3014603.0,
      "step": 1730
    },
    {
      "entropy": 6.536535358428955,
      "epoch": 0.13499319198599494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999632145608275,
      "loss": 6.2372,
      "mean_token_accuracy": 0.13129105269908906,
      "num_tokens": 3023170.0,
      "step": 1735
    },
    {
      "entropy": 6.278478670120239,
      "epoch": 0.1353822213577125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999627117037343,
      "loss": 5.9949,
      "mean_token_accuracy": 0.13947482481598855,
      "num_tokens": 3030565.0,
      "step": 1740
    },
    {
      "entropy": 6.4359362602233885,
      "epoch": 0.13577125072943008,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049996220543319,
      "loss": 6.2643,
      "mean_token_accuracy": 0.12935592159628867,
      "num_tokens": 3040124.0,
      "step": 1745
    },
    {
      "entropy": 6.372562122344971,
      "epoch": 0.13616028010114764,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999616957492025,
      "loss": 6.1289,
      "mean_token_accuracy": 0.13145363405346872,
      "num_tokens": 3048458.0,
      "step": 1750
    },
    {
      "entropy": 6.290291500091553,
      "epoch": 0.1365493094728652,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999611826517793,
      "loss": 6.103,
      "mean_token_accuracy": 0.1282423861324787,
      "num_tokens": 3057926.0,
      "step": 1755
    },
    {
      "entropy": 6.379631900787354,
      "epoch": 0.13693833884458276,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999606661409281,
      "loss": 6.1366,
      "mean_token_accuracy": 0.12733863443136215,
      "num_tokens": 3066683.0,
      "step": 1760
    },
    {
      "entropy": 6.335511159896851,
      "epoch": 0.13732736821630034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999601462166569,
      "loss": 6.2076,
      "mean_token_accuracy": 0.1272580251097679,
      "num_tokens": 3075117.0,
      "step": 1765
    },
    {
      "entropy": 6.486139631271362,
      "epoch": 0.1377163975880179,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999596228789736,
      "loss": 6.2551,
      "mean_token_accuracy": 0.12465462982654571,
      "num_tokens": 3084399.0,
      "step": 1770
    },
    {
      "entropy": 6.3006826877594,
      "epoch": 0.13810542695973546,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999590961278859,
      "loss": 6.0508,
      "mean_token_accuracy": 0.13070713505148887,
      "num_tokens": 3093030.0,
      "step": 1775
    },
    {
      "entropy": 6.385556221008301,
      "epoch": 0.13849445633145302,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999585659634021,
      "loss": 6.1168,
      "mean_token_accuracy": 0.12877569422125817,
      "num_tokens": 3101670.0,
      "step": 1780
    },
    {
      "entropy": 6.281736993789673,
      "epoch": 0.13888348570317058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999580323855301,
      "loss": 6.0969,
      "mean_token_accuracy": 0.1332916021347046,
      "num_tokens": 3109703.0,
      "step": 1785
    },
    {
      "entropy": 6.332530498504639,
      "epoch": 0.13927251507488816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999574953942781,
      "loss": 6.0721,
      "mean_token_accuracy": 0.1361841931939125,
      "num_tokens": 3118388.0,
      "step": 1790
    },
    {
      "entropy": 6.350433826446533,
      "epoch": 0.13966154444660572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999569549896541,
      "loss": 6.0765,
      "mean_token_accuracy": 0.13128025978803634,
      "num_tokens": 3126466.0,
      "step": 1795
    },
    {
      "entropy": 6.373475027084351,
      "epoch": 0.14005057381832328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999564111716665,
      "loss": 6.1227,
      "mean_token_accuracy": 0.13267524316906928,
      "num_tokens": 3135322.0,
      "step": 1800
    },
    {
      "entropy": 6.354067659378051,
      "epoch": 0.14043960319004084,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999558639403232,
      "loss": 6.1047,
      "mean_token_accuracy": 0.1384847082197666,
      "num_tokens": 3144498.0,
      "step": 1805
    },
    {
      "entropy": 6.28254337310791,
      "epoch": 0.1408286325617584,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999553132956328,
      "loss": 6.0917,
      "mean_token_accuracy": 0.12457187175750732,
      "num_tokens": 3153573.0,
      "step": 1810
    },
    {
      "entropy": 6.256527614593506,
      "epoch": 0.14121766193347599,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999547592376035,
      "loss": 6.0228,
      "mean_token_accuracy": 0.13926569893956184,
      "num_tokens": 3161705.0,
      "step": 1815
    },
    {
      "entropy": 6.382395887374878,
      "epoch": 0.14160669130519354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999542017662438,
      "loss": 6.1838,
      "mean_token_accuracy": 0.1349948912858963,
      "num_tokens": 3170068.0,
      "step": 1820
    },
    {
      "entropy": 6.34617772102356,
      "epoch": 0.1419957206769111,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999536408815623,
      "loss": 6.1935,
      "mean_token_accuracy": 0.12690433263778686,
      "num_tokens": 3178140.0,
      "step": 1825
    },
    {
      "entropy": 6.422201013565063,
      "epoch": 0.14238475004862866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999530765835672,
      "loss": 6.1579,
      "mean_token_accuracy": 0.1325029954314232,
      "num_tokens": 3186518.0,
      "step": 1830
    },
    {
      "entropy": 6.289926528930664,
      "epoch": 0.14277377942034625,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999525088722673,
      "loss": 6.1548,
      "mean_token_accuracy": 0.1283407174050808,
      "num_tokens": 3195246.0,
      "step": 1835
    },
    {
      "entropy": 6.444282150268554,
      "epoch": 0.1431628087920638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999519377476711,
      "loss": 6.1779,
      "mean_token_accuracy": 0.12667956873774527,
      "num_tokens": 3203946.0,
      "step": 1840
    },
    {
      "entropy": 6.306654262542724,
      "epoch": 0.14355183816378136,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999513632097872,
      "loss": 6.1739,
      "mean_token_accuracy": 0.13087522089481354,
      "num_tokens": 3212208.0,
      "step": 1845
    },
    {
      "entropy": 6.336764192581176,
      "epoch": 0.14394086753549892,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999507852586245,
      "loss": 6.1172,
      "mean_token_accuracy": 0.1314890868961811,
      "num_tokens": 3221572.0,
      "step": 1850
    },
    {
      "entropy": 6.424954843521118,
      "epoch": 0.14432989690721648,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999502038941917,
      "loss": 6.2263,
      "mean_token_accuracy": 0.12041985988616943,
      "num_tokens": 3230898.0,
      "step": 1855
    },
    {
      "entropy": 6.333400774002075,
      "epoch": 0.14471892627893407,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999496191164975,
      "loss": 6.2051,
      "mean_token_accuracy": 0.12520071864128113,
      "num_tokens": 3240524.0,
      "step": 1860
    },
    {
      "entropy": 6.29474139213562,
      "epoch": 0.14510795565065163,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499949030925551,
      "loss": 5.9728,
      "mean_token_accuracy": 0.13799916058778763,
      "num_tokens": 3248864.0,
      "step": 1865
    },
    {
      "entropy": 6.325572156906128,
      "epoch": 0.14549698502236919,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999484393213609,
      "loss": 6.1491,
      "mean_token_accuracy": 0.13214721530675888,
      "num_tokens": 3256823.0,
      "step": 1870
    },
    {
      "entropy": 6.3230939388275145,
      "epoch": 0.14588601439408674,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999478443039363,
      "loss": 6.0723,
      "mean_token_accuracy": 0.13165922313928605,
      "num_tokens": 3265621.0,
      "step": 1875
    },
    {
      "entropy": 6.260721683502197,
      "epoch": 0.14627504376580433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999472458732862,
      "loss": 6.0962,
      "mean_token_accuracy": 0.13295478150248527,
      "num_tokens": 3273515.0,
      "step": 1880
    },
    {
      "entropy": 6.414492464065551,
      "epoch": 0.1466640731375219,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999466440294198,
      "loss": 6.1689,
      "mean_token_accuracy": 0.12950320392847062,
      "num_tokens": 3282283.0,
      "step": 1885
    },
    {
      "entropy": 6.302510833740234,
      "epoch": 0.14705310250923945,
      "grad_norm": 1.0,
      "learning_rate": 0.000499946038772346,
      "loss": 6.1611,
      "mean_token_accuracy": 0.13020878359675409,
      "num_tokens": 3290792.0,
      "step": 1890
    },
    {
      "entropy": 6.255168008804321,
      "epoch": 0.147442131880957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999454301020741,
      "loss": 6.0204,
      "mean_token_accuracy": 0.13267397582530976,
      "num_tokens": 3299430.0,
      "step": 1895
    },
    {
      "entropy": 6.261677408218384,
      "epoch": 0.14783116125267456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999448180186133,
      "loss": 6.0808,
      "mean_token_accuracy": 0.13431849405169488,
      "num_tokens": 3307949.0,
      "step": 1900
    },
    {
      "entropy": 6.340135097503662,
      "epoch": 0.14822019062439215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499944202521973,
      "loss": 6.1246,
      "mean_token_accuracy": 0.13576969206333162,
      "num_tokens": 3316900.0,
      "step": 1905
    },
    {
      "entropy": 6.295166778564453,
      "epoch": 0.1486092199961097,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999435836121624,
      "loss": 6.0446,
      "mean_token_accuracy": 0.13577630817890168,
      "num_tokens": 3325656.0,
      "step": 1910
    },
    {
      "entropy": 6.309445333480835,
      "epoch": 0.14899824936782727,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499942961289191,
      "loss": 6.2644,
      "mean_token_accuracy": 0.12280140817165375,
      "num_tokens": 3334219.0,
      "step": 1915
    },
    {
      "entropy": 6.317688608169556,
      "epoch": 0.14938727873954483,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999423355530683,
      "loss": 6.0314,
      "mean_token_accuracy": 0.13022207394242286,
      "num_tokens": 3343637.0,
      "step": 1920
    },
    {
      "entropy": 6.314282560348511,
      "epoch": 0.1497763081112624,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999417064038035,
      "loss": 6.1686,
      "mean_token_accuracy": 0.13118190541863442,
      "num_tokens": 3352624.0,
      "step": 1925
    },
    {
      "entropy": 6.262926197052002,
      "epoch": 0.15016533748297997,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999410738414065,
      "loss": 6.0802,
      "mean_token_accuracy": 0.1389567419886589,
      "num_tokens": 3361394.0,
      "step": 1930
    },
    {
      "entropy": 6.377456378936768,
      "epoch": 0.15055436685469753,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999404378658866,
      "loss": 6.1348,
      "mean_token_accuracy": 0.13395721539855004,
      "num_tokens": 3369406.0,
      "step": 1935
    },
    {
      "entropy": 6.3417693138122555,
      "epoch": 0.1509433962264151,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999397984772536,
      "loss": 6.0852,
      "mean_token_accuracy": 0.13622010201215745,
      "num_tokens": 3377737.0,
      "step": 1940
    },
    {
      "entropy": 6.2485908508300785,
      "epoch": 0.15133242559813265,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999391556755172,
      "loss": 6.1389,
      "mean_token_accuracy": 0.12751591503620147,
      "num_tokens": 3385514.0,
      "step": 1945
    },
    {
      "entropy": 6.394997596740723,
      "epoch": 0.15172145496985023,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999385094606872,
      "loss": 6.0537,
      "mean_token_accuracy": 0.13509045764803887,
      "num_tokens": 3394500.0,
      "step": 1950
    },
    {
      "entropy": 6.220421218872071,
      "epoch": 0.1521104843415678,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999378598327733,
      "loss": 6.0306,
      "mean_token_accuracy": 0.13421875312924386,
      "num_tokens": 3403289.0,
      "step": 1955
    },
    {
      "entropy": 6.218709182739258,
      "epoch": 0.15249951371328535,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999372067917854,
      "loss": 5.9569,
      "mean_token_accuracy": 0.14868551939725877,
      "num_tokens": 3410892.0,
      "step": 1960
    },
    {
      "entropy": 6.162004232406616,
      "epoch": 0.1528885430850029,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999365503377335,
      "loss": 6.061,
      "mean_token_accuracy": 0.13821410536766052,
      "num_tokens": 3420329.0,
      "step": 1965
    },
    {
      "entropy": 6.333985614776611,
      "epoch": 0.1532775724567205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999358904706273,
      "loss": 6.0505,
      "mean_token_accuracy": 0.1318977616727352,
      "num_tokens": 3429429.0,
      "step": 1970
    },
    {
      "entropy": 6.335431814193726,
      "epoch": 0.15366660182843805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999352271904771,
      "loss": 6.1639,
      "mean_token_accuracy": 0.1311632439494133,
      "num_tokens": 3438198.0,
      "step": 1975
    },
    {
      "entropy": 6.457853221893311,
      "epoch": 0.1540556312001556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999345604972928,
      "loss": 6.2972,
      "mean_token_accuracy": 0.1280217707157135,
      "num_tokens": 3447732.0,
      "step": 1980
    },
    {
      "entropy": 6.3014496803283695,
      "epoch": 0.15444466057187317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999338903910846,
      "loss": 6.0717,
      "mean_token_accuracy": 0.13846041038632392,
      "num_tokens": 3456392.0,
      "step": 1985
    },
    {
      "entropy": 6.261079359054565,
      "epoch": 0.15483368994359073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999332168718626,
      "loss": 6.1141,
      "mean_token_accuracy": 0.13328127637505532,
      "num_tokens": 3464473.0,
      "step": 1990
    },
    {
      "entropy": 6.2995849609375,
      "epoch": 0.15522271931530832,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999325399396371,
      "loss": 6.0817,
      "mean_token_accuracy": 0.13139236569404603,
      "num_tokens": 3473740.0,
      "step": 1995
    },
    {
      "entropy": 6.162753677368164,
      "epoch": 0.15561174868702587,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999318595944183,
      "loss": 5.985,
      "mean_token_accuracy": 0.1351643517613411,
      "num_tokens": 3482249.0,
      "step": 2000
    },
    {
      "entropy": 6.419935607910157,
      "epoch": 0.15600077805874343,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999311758362166,
      "loss": 6.2085,
      "mean_token_accuracy": 0.12761779576539994,
      "num_tokens": 3490753.0,
      "step": 2005
    },
    {
      "entropy": 6.390731334686279,
      "epoch": 0.156389807430461,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999304886650422,
      "loss": 6.1566,
      "mean_token_accuracy": 0.13517940416932106,
      "num_tokens": 3499373.0,
      "step": 2010
    },
    {
      "entropy": 6.3211747169494625,
      "epoch": 0.15677883680217858,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999297980809058,
      "loss": 6.1308,
      "mean_token_accuracy": 0.12607129365205766,
      "num_tokens": 3508344.0,
      "step": 2015
    },
    {
      "entropy": 6.359952831268311,
      "epoch": 0.15716786617389614,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999291040838176,
      "loss": 6.0979,
      "mean_token_accuracy": 0.1342169925570488,
      "num_tokens": 3517009.0,
      "step": 2020
    },
    {
      "entropy": 6.2388135433197025,
      "epoch": 0.1575568955456137,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999284066737884,
      "loss": 6.1366,
      "mean_token_accuracy": 0.13406251445412637,
      "num_tokens": 3525989.0,
      "step": 2025
    },
    {
      "entropy": 6.306530141830445,
      "epoch": 0.15794592491733125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999277058508285,
      "loss": 6.0173,
      "mean_token_accuracy": 0.13108256459236145,
      "num_tokens": 3533947.0,
      "step": 2030
    },
    {
      "entropy": 6.262210369110107,
      "epoch": 0.1583349542890488,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999270016149486,
      "loss": 6.0749,
      "mean_token_accuracy": 0.1352415882050991,
      "num_tokens": 3541974.0,
      "step": 2035
    },
    {
      "entropy": 6.268887376785278,
      "epoch": 0.1587239836607664,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999262939661596,
      "loss": 6.0026,
      "mean_token_accuracy": 0.1375070445239544,
      "num_tokens": 3550436.0,
      "step": 2040
    },
    {
      "entropy": 6.26065878868103,
      "epoch": 0.15911301303248396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999255829044721,
      "loss": 6.0238,
      "mean_token_accuracy": 0.13955531716346742,
      "num_tokens": 3558283.0,
      "step": 2045
    },
    {
      "entropy": 6.194809818267823,
      "epoch": 0.15950204240420152,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999248684298968,
      "loss": 5.939,
      "mean_token_accuracy": 0.1412186801433563,
      "num_tokens": 3566968.0,
      "step": 2050
    },
    {
      "entropy": 6.244066286087036,
      "epoch": 0.15989107177591907,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999241505424447,
      "loss": 6.0625,
      "mean_token_accuracy": 0.12982572838664055,
      "num_tokens": 3575580.0,
      "step": 2055
    },
    {
      "entropy": 6.259646034240722,
      "epoch": 0.16028010114763663,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999234292421265,
      "loss": 6.1312,
      "mean_token_accuracy": 0.13408922925591468,
      "num_tokens": 3584389.0,
      "step": 2060
    },
    {
      "entropy": 6.286141872406006,
      "epoch": 0.16066913051935422,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999227045289535,
      "loss": 5.9685,
      "mean_token_accuracy": 0.13707195296883584,
      "num_tokens": 3593009.0,
      "step": 2065
    },
    {
      "entropy": 6.142366409301758,
      "epoch": 0.16105815989107178,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999219764029363,
      "loss": 6.0046,
      "mean_token_accuracy": 0.13540613800287246,
      "num_tokens": 3601914.0,
      "step": 2070
    },
    {
      "entropy": 6.207937431335449,
      "epoch": 0.16144718926278934,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999212448640861,
      "loss": 5.9673,
      "mean_token_accuracy": 0.13867463171482086,
      "num_tokens": 3609883.0,
      "step": 2075
    },
    {
      "entropy": 6.189123487472534,
      "epoch": 0.1618362186345069,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999205099124141,
      "loss": 5.9265,
      "mean_token_accuracy": 0.1397370256483555,
      "num_tokens": 3618103.0,
      "step": 2080
    },
    {
      "entropy": 6.399334859848023,
      "epoch": 0.16222524800622448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999197715479313,
      "loss": 6.1937,
      "mean_token_accuracy": 0.1316985122859478,
      "num_tokens": 3625861.0,
      "step": 2085
    },
    {
      "entropy": 6.1656725883483885,
      "epoch": 0.16261427737794204,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999190297706489,
      "loss": 5.9726,
      "mean_token_accuracy": 0.13532934188842774,
      "num_tokens": 3634498.0,
      "step": 2090
    },
    {
      "entropy": 6.1408239841461185,
      "epoch": 0.1630033067496596,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999182845805783,
      "loss": 5.9117,
      "mean_token_accuracy": 0.13378374204039573,
      "num_tokens": 3644186.0,
      "step": 2095
    },
    {
      "entropy": 6.151297092437744,
      "epoch": 0.16339233612137716,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999175359777306,
      "loss": 6.0603,
      "mean_token_accuracy": 0.13409929126501083,
      "num_tokens": 3652831.0,
      "step": 2100
    },
    {
      "entropy": 6.346699666976929,
      "epoch": 0.16378136549309472,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999167839621174,
      "loss": 5.9205,
      "mean_token_accuracy": 0.14053727090358734,
      "num_tokens": 3661230.0,
      "step": 2105
    },
    {
      "entropy": 6.19318904876709,
      "epoch": 0.1641703948648123,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999160285337501,
      "loss": 6.0607,
      "mean_token_accuracy": 0.1370956338942051,
      "num_tokens": 3669250.0,
      "step": 2110
    },
    {
      "entropy": 6.26489109992981,
      "epoch": 0.16455942423652986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999152696926399,
      "loss": 6.0286,
      "mean_token_accuracy": 0.138458015024662,
      "num_tokens": 3677568.0,
      "step": 2115
    },
    {
      "entropy": 6.253146648406982,
      "epoch": 0.16494845360824742,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999145074387985,
      "loss": 6.0233,
      "mean_token_accuracy": 0.1416604608297348,
      "num_tokens": 3685488.0,
      "step": 2120
    },
    {
      "entropy": 6.180020809173584,
      "epoch": 0.16533748297996498,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999137417722374,
      "loss": 6.002,
      "mean_token_accuracy": 0.13473087772727013,
      "num_tokens": 3694357.0,
      "step": 2125
    },
    {
      "entropy": 6.316441535949707,
      "epoch": 0.16572651235168256,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999129726929684,
      "loss": 6.0723,
      "mean_token_accuracy": 0.1360612131655216,
      "num_tokens": 3703173.0,
      "step": 2130
    },
    {
      "entropy": 6.205403232574463,
      "epoch": 0.16611554172340012,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999122002010029,
      "loss": 6.0431,
      "mean_token_accuracy": 0.14090028777718544,
      "num_tokens": 3712070.0,
      "step": 2135
    },
    {
      "entropy": 6.206867980957031,
      "epoch": 0.16650457109511768,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999114242963527,
      "loss": 5.9404,
      "mean_token_accuracy": 0.14348391592502593,
      "num_tokens": 3720245.0,
      "step": 2140
    },
    {
      "entropy": 6.18376841545105,
      "epoch": 0.16689360046683524,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999106449790298,
      "loss": 6.0492,
      "mean_token_accuracy": 0.137314972281456,
      "num_tokens": 3730064.0,
      "step": 2145
    },
    {
      "entropy": 6.136141538619995,
      "epoch": 0.1672826298385528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999098622490458,
      "loss": 5.9864,
      "mean_token_accuracy": 0.14003208428621292,
      "num_tokens": 3738741.0,
      "step": 2150
    },
    {
      "entropy": 6.289495420455933,
      "epoch": 0.16767165921027039,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999090761064125,
      "loss": 6.0533,
      "mean_token_accuracy": 0.13532209545373916,
      "num_tokens": 3747472.0,
      "step": 2155
    },
    {
      "entropy": 6.240089464187622,
      "epoch": 0.16806068858198794,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499908286551142,
      "loss": 6.0641,
      "mean_token_accuracy": 0.13511611968278886,
      "num_tokens": 3756643.0,
      "step": 2160
    },
    {
      "entropy": 6.315414953231811,
      "epoch": 0.1684497179537055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999074935832463,
      "loss": 6.1853,
      "mean_token_accuracy": 0.12947507873177527,
      "num_tokens": 3765490.0,
      "step": 2165
    },
    {
      "entropy": 6.2349205493927,
      "epoch": 0.16883874732542306,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999066972027373,
      "loss": 5.938,
      "mean_token_accuracy": 0.1387890338897705,
      "num_tokens": 3773063.0,
      "step": 2170
    },
    {
      "entropy": 6.261452722549438,
      "epoch": 0.16922777669714065,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999058974096271,
      "loss": 6.1062,
      "mean_token_accuracy": 0.1365394599735737,
      "num_tokens": 3781763.0,
      "step": 2175
    },
    {
      "entropy": 6.2590172290802,
      "epoch": 0.1696168060688582,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499905094203928,
      "loss": 6.0982,
      "mean_token_accuracy": 0.1321517176926136,
      "num_tokens": 3790582.0,
      "step": 2180
    },
    {
      "entropy": 6.202856731414795,
      "epoch": 0.17000583544057576,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499904287585652,
      "loss": 5.9206,
      "mean_token_accuracy": 0.13618794158101083,
      "num_tokens": 3798745.0,
      "step": 2185
    },
    {
      "entropy": 6.172497653961182,
      "epoch": 0.17039486481229332,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999034775548113,
      "loss": 6.0095,
      "mean_token_accuracy": 0.1356318287551403,
      "num_tokens": 3806901.0,
      "step": 2190
    },
    {
      "entropy": 6.2153675079345705,
      "epoch": 0.17078389418401088,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999026641114185,
      "loss": 5.9479,
      "mean_token_accuracy": 0.14665559828281402,
      "num_tokens": 3815742.0,
      "step": 2195
    },
    {
      "entropy": 6.144705820083618,
      "epoch": 0.17117292355572847,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999018472554857,
      "loss": 6.0388,
      "mean_token_accuracy": 0.13324448689818383,
      "num_tokens": 3824468.0,
      "step": 2200
    },
    {
      "entropy": 6.208492565155029,
      "epoch": 0.17156195292744603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999010269870253,
      "loss": 5.9937,
      "mean_token_accuracy": 0.13999679386615754,
      "num_tokens": 3832834.0,
      "step": 2205
    },
    {
      "entropy": 6.242017364501953,
      "epoch": 0.17195098229916359,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999002033060498,
      "loss": 6.0525,
      "mean_token_accuracy": 0.13752312287688256,
      "num_tokens": 3841780.0,
      "step": 2210
    },
    {
      "entropy": 6.24888129234314,
      "epoch": 0.17234001167088114,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998993762125716,
      "loss": 6.1355,
      "mean_token_accuracy": 0.12495578080415726,
      "num_tokens": 3849840.0,
      "step": 2215
    },
    {
      "entropy": 6.293956756591797,
      "epoch": 0.17272904104259873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998985457066034,
      "loss": 6.079,
      "mean_token_accuracy": 0.13282646834850312,
      "num_tokens": 3858204.0,
      "step": 2220
    },
    {
      "entropy": 6.2079095363616945,
      "epoch": 0.1731180704143163,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998977117881576,
      "loss": 6.0237,
      "mean_token_accuracy": 0.14005702137947082,
      "num_tokens": 3866615.0,
      "step": 2225
    },
    {
      "entropy": 6.221766662597656,
      "epoch": 0.17350709978603385,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998968744572472,
      "loss": 6.0559,
      "mean_token_accuracy": 0.14006396681070327,
      "num_tokens": 3876003.0,
      "step": 2230
    },
    {
      "entropy": 6.193023920059204,
      "epoch": 0.1738961291577514,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998960337138845,
      "loss": 5.9727,
      "mean_token_accuracy": 0.13838636204600335,
      "num_tokens": 3884362.0,
      "step": 2235
    },
    {
      "entropy": 6.160814571380615,
      "epoch": 0.17428515852946896,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998951895580826,
      "loss": 5.9274,
      "mean_token_accuracy": 0.13808080181479454,
      "num_tokens": 3893177.0,
      "step": 2240
    },
    {
      "entropy": 6.1626489639282225,
      "epoch": 0.17467418790118655,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998943419898542,
      "loss": 5.9864,
      "mean_token_accuracy": 0.14280490577220917,
      "num_tokens": 3903006.0,
      "step": 2245
    },
    {
      "entropy": 6.204087400436402,
      "epoch": 0.1750632172729041,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998934910092119,
      "loss": 5.8424,
      "mean_token_accuracy": 0.1409443661570549,
      "num_tokens": 3912256.0,
      "step": 2250
    },
    {
      "entropy": 6.134327507019043,
      "epoch": 0.17545224664462167,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998926366161689,
      "loss": 6.0321,
      "mean_token_accuracy": 0.13451322615146638,
      "num_tokens": 3921121.0,
      "step": 2255
    },
    {
      "entropy": 6.185262537002563,
      "epoch": 0.17584127601633923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998917788107383,
      "loss": 5.9843,
      "mean_token_accuracy": 0.13937503695487977,
      "num_tokens": 3929348.0,
      "step": 2260
    },
    {
      "entropy": 6.260100173950195,
      "epoch": 0.17623030538805678,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998909175929326,
      "loss": 6.1355,
      "mean_token_accuracy": 0.13460659310221673,
      "num_tokens": 3938073.0,
      "step": 2265
    },
    {
      "entropy": 6.28374981880188,
      "epoch": 0.17661933475977437,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998900529627653,
      "loss": 6.1158,
      "mean_token_accuracy": 0.13738646805286409,
      "num_tokens": 3946641.0,
      "step": 2270
    },
    {
      "entropy": 6.256800127029419,
      "epoch": 0.17700836413149193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998891849202495,
      "loss": 5.9659,
      "mean_token_accuracy": 0.1371592253446579,
      "num_tokens": 3954586.0,
      "step": 2275
    },
    {
      "entropy": 6.119969654083252,
      "epoch": 0.1773973935032095,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499888313465398,
      "loss": 6.0389,
      "mean_token_accuracy": 0.14258864596486093,
      "num_tokens": 3963745.0,
      "step": 2280
    },
    {
      "entropy": 6.1104155540466305,
      "epoch": 0.17778642287492705,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998874385982247,
      "loss": 5.8596,
      "mean_token_accuracy": 0.1476686991751194,
      "num_tokens": 3972452.0,
      "step": 2285
    },
    {
      "entropy": 6.156691980361939,
      "epoch": 0.17817545224664463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998865603187421,
      "loss": 5.9683,
      "mean_token_accuracy": 0.13907164037227632,
      "num_tokens": 3980770.0,
      "step": 2290
    },
    {
      "entropy": 6.192203378677368,
      "epoch": 0.1785644816183622,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499885678626964,
      "loss": 5.9986,
      "mean_token_accuracy": 0.13721442893147467,
      "num_tokens": 3989545.0,
      "step": 2295
    },
    {
      "entropy": 6.25260329246521,
      "epoch": 0.17895351099007975,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998847935229038,
      "loss": 5.9638,
      "mean_token_accuracy": 0.13558776378631593,
      "num_tokens": 3998034.0,
      "step": 2300
    },
    {
      "entropy": 6.122658395767212,
      "epoch": 0.1793425403617973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998839050065746,
      "loss": 5.9926,
      "mean_token_accuracy": 0.1411259353160858,
      "num_tokens": 4007110.0,
      "step": 2305
    },
    {
      "entropy": 6.245479536056519,
      "epoch": 0.17973156973351487,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998830130779902,
      "loss": 6.0628,
      "mean_token_accuracy": 0.13011873438954352,
      "num_tokens": 4015930.0,
      "step": 2310
    },
    {
      "entropy": 6.104860258102417,
      "epoch": 0.18012059910523245,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499882117737164,
      "loss": 5.8558,
      "mean_token_accuracy": 0.1457628294825554,
      "num_tokens": 4024510.0,
      "step": 2315
    },
    {
      "entropy": 6.106551742553711,
      "epoch": 0.18050962847695,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998812189841096,
      "loss": 5.9437,
      "mean_token_accuracy": 0.14204365611076356,
      "num_tokens": 4033714.0,
      "step": 2320
    },
    {
      "entropy": 6.316892910003662,
      "epoch": 0.18089865784866757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998803168188407,
      "loss": 6.1612,
      "mean_token_accuracy": 0.13279995173215867,
      "num_tokens": 4042376.0,
      "step": 2325
    },
    {
      "entropy": 6.147280216217041,
      "epoch": 0.18128768722038513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998794112413708,
      "loss": 5.925,
      "mean_token_accuracy": 0.13952778577804564,
      "num_tokens": 4050521.0,
      "step": 2330
    },
    {
      "entropy": 6.251437854766846,
      "epoch": 0.18167671659210272,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998785022517137,
      "loss": 6.0965,
      "mean_token_accuracy": 0.13085465058684348,
      "num_tokens": 4059384.0,
      "step": 2335
    },
    {
      "entropy": 6.1695960521697994,
      "epoch": 0.18206574596382027,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998775898498835,
      "loss": 5.9663,
      "mean_token_accuracy": 0.14308709353208543,
      "num_tokens": 4067655.0,
      "step": 2340
    },
    {
      "entropy": 6.172775983810425,
      "epoch": 0.18245477533553783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998766740358936,
      "loss": 5.9634,
      "mean_token_accuracy": 0.14366720467805863,
      "num_tokens": 4075828.0,
      "step": 2345
    },
    {
      "entropy": 6.232926893234253,
      "epoch": 0.1828438047072554,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998757548097582,
      "loss": 5.9226,
      "mean_token_accuracy": 0.1397922396659851,
      "num_tokens": 4083796.0,
      "step": 2350
    },
    {
      "entropy": 6.120053386688232,
      "epoch": 0.18323283407897295,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998748321714911,
      "loss": 6.0329,
      "mean_token_accuracy": 0.14252227693796157,
      "num_tokens": 4092981.0,
      "step": 2355
    },
    {
      "entropy": 6.174644327163696,
      "epoch": 0.18362186345069054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998739061211065,
      "loss": 5.879,
      "mean_token_accuracy": 0.13968045711517335,
      "num_tokens": 4101363.0,
      "step": 2360
    },
    {
      "entropy": 6.13595757484436,
      "epoch": 0.1840108928224081,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998729766586181,
      "loss": 5.9492,
      "mean_token_accuracy": 0.1380427822470665,
      "num_tokens": 4110164.0,
      "step": 2365
    },
    {
      "entropy": 6.248952150344849,
      "epoch": 0.18439992219412565,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998720437840403,
      "loss": 6.0636,
      "mean_token_accuracy": 0.1388567715883255,
      "num_tokens": 4119098.0,
      "step": 2370
    },
    {
      "entropy": 6.176186418533325,
      "epoch": 0.1847889515658432,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998711074973871,
      "loss": 5.9613,
      "mean_token_accuracy": 0.14075352251529694,
      "num_tokens": 4127522.0,
      "step": 2375
    },
    {
      "entropy": 6.1555253028869625,
      "epoch": 0.1851779809375608,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998701677986728,
      "loss": 5.9933,
      "mean_token_accuracy": 0.1459673412144184,
      "num_tokens": 4136311.0,
      "step": 2380
    },
    {
      "entropy": 6.152952861785889,
      "epoch": 0.18556701030927836,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998692246879116,
      "loss": 5.9173,
      "mean_token_accuracy": 0.14190907552838325,
      "num_tokens": 4145160.0,
      "step": 2385
    },
    {
      "entropy": 6.123033618927002,
      "epoch": 0.18595603968099592,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998682781651178,
      "loss": 5.8841,
      "mean_token_accuracy": 0.1437641464173794,
      "num_tokens": 4153921.0,
      "step": 2390
    },
    {
      "entropy": 6.115445232391357,
      "epoch": 0.18634506905271347,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998673282303059,
      "loss": 5.9681,
      "mean_token_accuracy": 0.1352468617260456,
      "num_tokens": 4162580.0,
      "step": 2395
    },
    {
      "entropy": 6.182539653778076,
      "epoch": 0.18673409842443103,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998663748834901,
      "loss": 5.9663,
      "mean_token_accuracy": 0.14548829048871995,
      "num_tokens": 4170623.0,
      "step": 2400
    },
    {
      "entropy": 6.110377979278565,
      "epoch": 0.18712312779614862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998654181246851,
      "loss": 5.9514,
      "mean_token_accuracy": 0.14497750550508498,
      "num_tokens": 4178724.0,
      "step": 2405
    },
    {
      "entropy": 6.137342739105224,
      "epoch": 0.18751215716786618,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998644579539052,
      "loss": 5.9613,
      "mean_token_accuracy": 0.1416378900408745,
      "num_tokens": 4187773.0,
      "step": 2410
    },
    {
      "entropy": 6.170235538482666,
      "epoch": 0.18790118653958374,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499863494371165,
      "loss": 5.9529,
      "mean_token_accuracy": 0.13757317140698433,
      "num_tokens": 4196628.0,
      "step": 2415
    },
    {
      "entropy": 6.15802321434021,
      "epoch": 0.1882902159113013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998625273764793,
      "loss": 5.9779,
      "mean_token_accuracy": 0.1390997901558876,
      "num_tokens": 4205090.0,
      "step": 2420
    },
    {
      "entropy": 6.209850788116455,
      "epoch": 0.18867924528301888,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998615569698626,
      "loss": 5.8944,
      "mean_token_accuracy": 0.1414852797985077,
      "num_tokens": 4212757.0,
      "step": 2425
    },
    {
      "entropy": 6.018552446365357,
      "epoch": 0.18906827465473644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998605831513296,
      "loss": 5.9541,
      "mean_token_accuracy": 0.14190278872847556,
      "num_tokens": 4221556.0,
      "step": 2430
    },
    {
      "entropy": 6.107316589355468,
      "epoch": 0.189457304026454,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998596059208953,
      "loss": 5.9427,
      "mean_token_accuracy": 0.13826509863138198,
      "num_tokens": 4231524.0,
      "step": 2435
    },
    {
      "entropy": 6.196195983886719,
      "epoch": 0.18984633339817156,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998586252785743,
      "loss": 5.9895,
      "mean_token_accuracy": 0.1437416195869446,
      "num_tokens": 4240229.0,
      "step": 2440
    },
    {
      "entropy": 6.146422052383423,
      "epoch": 0.19023536276988912,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998576412243816,
      "loss": 5.8743,
      "mean_token_accuracy": 0.14444278329610824,
      "num_tokens": 4248053.0,
      "step": 2445
    },
    {
      "entropy": 6.164689350128174,
      "epoch": 0.1906243921416067,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998566537583321,
      "loss": 5.9185,
      "mean_token_accuracy": 0.14325421303510666,
      "num_tokens": 4256019.0,
      "step": 2450
    },
    {
      "entropy": 6.0683635711669925,
      "epoch": 0.19101342151332426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998556628804408,
      "loss": 5.9053,
      "mean_token_accuracy": 0.14163314551115036,
      "num_tokens": 4263805.0,
      "step": 2455
    },
    {
      "entropy": 6.10060682296753,
      "epoch": 0.19140245088504182,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998546685907225,
      "loss": 5.9544,
      "mean_token_accuracy": 0.13736186251044274,
      "num_tokens": 4272712.0,
      "step": 2460
    },
    {
      "entropy": 6.20326189994812,
      "epoch": 0.19179148025675938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998536708891927,
      "loss": 5.8959,
      "mean_token_accuracy": 0.1452920749783516,
      "num_tokens": 4281222.0,
      "step": 2465
    },
    {
      "entropy": 6.094033193588257,
      "epoch": 0.19218050962847694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998526697758663,
      "loss": 6.0001,
      "mean_token_accuracy": 0.13505400642752646,
      "num_tokens": 4290558.0,
      "step": 2470
    },
    {
      "entropy": 6.264354085922241,
      "epoch": 0.19256953900019452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998516652507585,
      "loss": 6.0245,
      "mean_token_accuracy": 0.13424147367477418,
      "num_tokens": 4299684.0,
      "step": 2475
    },
    {
      "entropy": 6.101936912536621,
      "epoch": 0.19295856837191208,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998506573138846,
      "loss": 5.9608,
      "mean_token_accuracy": 0.14035997539758682,
      "num_tokens": 4308938.0,
      "step": 2480
    },
    {
      "entropy": 6.069972848892212,
      "epoch": 0.19334759774362964,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998496459652598,
      "loss": 5.828,
      "mean_token_accuracy": 0.147909976541996,
      "num_tokens": 4317646.0,
      "step": 2485
    },
    {
      "entropy": 6.097158575057984,
      "epoch": 0.1937366271153472,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998486312048995,
      "loss": 5.9509,
      "mean_token_accuracy": 0.14012686312198638,
      "num_tokens": 4326876.0,
      "step": 2490
    },
    {
      "entropy": 6.119687271118164,
      "epoch": 0.19412565648706478,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998476130328193,
      "loss": 5.9374,
      "mean_token_accuracy": 0.14259177893400193,
      "num_tokens": 4335391.0,
      "step": 2495
    },
    {
      "entropy": 6.211879730224609,
      "epoch": 0.19451468585878234,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998465914490343,
      "loss": 5.9104,
      "mean_token_accuracy": 0.1425481393933296,
      "num_tokens": 4343253.0,
      "step": 2500
    },
    {
      "entropy": 5.971693420410157,
      "epoch": 0.1949037152304999,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049984556645356,
      "loss": 5.8904,
      "mean_token_accuracy": 0.13843516409397125,
      "num_tokens": 4352161.0,
      "step": 2505
    },
    {
      "entropy": 6.155507898330688,
      "epoch": 0.19529274460221746,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998445380464124,
      "loss": 5.853,
      "mean_token_accuracy": 0.1473579704761505,
      "num_tokens": 4360414.0,
      "step": 2510
    },
    {
      "entropy": 6.0863183498382565,
      "epoch": 0.19568177397393502,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998435062276066,
      "loss": 5.8366,
      "mean_token_accuracy": 0.15001533925533295,
      "num_tokens": 4369375.0,
      "step": 2515
    },
    {
      "entropy": 5.990550470352173,
      "epoch": 0.1960708033456526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998424709971586,
      "loss": 5.8722,
      "mean_token_accuracy": 0.14208223000168801,
      "num_tokens": 4378638.0,
      "step": 2520
    },
    {
      "entropy": 6.178989458084106,
      "epoch": 0.19645983271737016,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998414323550839,
      "loss": 5.8595,
      "mean_token_accuracy": 0.14671409651637077,
      "num_tokens": 4388262.0,
      "step": 2525
    },
    {
      "entropy": 6.066281509399414,
      "epoch": 0.19684886208908772,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998403903013984,
      "loss": 5.864,
      "mean_token_accuracy": 0.14327401518821717,
      "num_tokens": 4396894.0,
      "step": 2530
    },
    {
      "entropy": 6.139949607849121,
      "epoch": 0.19723789146080528,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998393448361179,
      "loss": 5.9885,
      "mean_token_accuracy": 0.14116159975528716,
      "num_tokens": 4405455.0,
      "step": 2535
    },
    {
      "entropy": 6.167979383468628,
      "epoch": 0.19762692083252287,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499838295959258,
      "loss": 5.8999,
      "mean_token_accuracy": 0.1390222877264023,
      "num_tokens": 4414355.0,
      "step": 2540
    },
    {
      "entropy": 5.997927045822143,
      "epoch": 0.19801595020424043,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499837243670835,
      "loss": 5.7913,
      "mean_token_accuracy": 0.15371408313512802,
      "num_tokens": 4422305.0,
      "step": 2545
    },
    {
      "entropy": 6.17957763671875,
      "epoch": 0.19840497957595798,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998361879708646,
      "loss": 5.9198,
      "mean_token_accuracy": 0.1427523523569107,
      "num_tokens": 4430813.0,
      "step": 2550
    },
    {
      "entropy": 6.099480628967285,
      "epoch": 0.19879400894767554,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499835128859363,
      "loss": 5.8828,
      "mean_token_accuracy": 0.1461223766207695,
      "num_tokens": 4440229.0,
      "step": 2555
    },
    {
      "entropy": 6.00967698097229,
      "epoch": 0.1991830383193931,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998340663363461,
      "loss": 5.8994,
      "mean_token_accuracy": 0.14125050902366637,
      "num_tokens": 4449417.0,
      "step": 2560
    },
    {
      "entropy": 6.186902904510498,
      "epoch": 0.1995720676911107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998330004018301,
      "loss": 5.9456,
      "mean_token_accuracy": 0.13707190454006196,
      "num_tokens": 4457377.0,
      "step": 2565
    },
    {
      "entropy": 6.101526355743408,
      "epoch": 0.19996109706282825,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998319310558312,
      "loss": 5.8663,
      "mean_token_accuracy": 0.14043352231383324,
      "num_tokens": 4466839.0,
      "step": 2570
    },
    {
      "entropy": 6.076670742034912,
      "epoch": 0.2003501264345458,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998308582983657,
      "loss": 5.8854,
      "mean_token_accuracy": 0.14421407207846643,
      "num_tokens": 4476369.0,
      "step": 2575
    },
    {
      "entropy": 6.072381830215454,
      "epoch": 0.20073915580626336,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998297821294497,
      "loss": 5.9561,
      "mean_token_accuracy": 0.14137691259384155,
      "num_tokens": 4485609.0,
      "step": 2580
    },
    {
      "entropy": 6.131612968444824,
      "epoch": 0.20112818517798095,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998287025490995,
      "loss": 5.9162,
      "mean_token_accuracy": 0.14042819067835807,
      "num_tokens": 4494535.0,
      "step": 2585
    },
    {
      "entropy": 6.1383223056793215,
      "epoch": 0.2015172145496985,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998276195573317,
      "loss": 5.8955,
      "mean_token_accuracy": 0.13853249028325082,
      "num_tokens": 4503133.0,
      "step": 2590
    },
    {
      "entropy": 6.049847364425659,
      "epoch": 0.20190624392141607,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998265331541627,
      "loss": 5.7882,
      "mean_token_accuracy": 0.149039888381958,
      "num_tokens": 4511465.0,
      "step": 2595
    },
    {
      "entropy": 6.050959205627441,
      "epoch": 0.20229527329313363,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998254433396088,
      "loss": 5.8551,
      "mean_token_accuracy": 0.14632439315319062,
      "num_tokens": 4520177.0,
      "step": 2600
    },
    {
      "entropy": 6.156892967224121,
      "epoch": 0.20268430266485118,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998243501136867,
      "loss": 6.0153,
      "mean_token_accuracy": 0.14031418338418006,
      "num_tokens": 4528370.0,
      "step": 2605
    },
    {
      "entropy": 6.219958877563476,
      "epoch": 0.20307333203656877,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998232534764129,
      "loss": 6.0333,
      "mean_token_accuracy": 0.13372237458825112,
      "num_tokens": 4538258.0,
      "step": 2610
    },
    {
      "entropy": 6.0912260055542,
      "epoch": 0.20346236140828633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998221534278041,
      "loss": 5.9243,
      "mean_token_accuracy": 0.14148016050457954,
      "num_tokens": 4547300.0,
      "step": 2615
    },
    {
      "entropy": 6.163697814941406,
      "epoch": 0.2038513907800039,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998210499678769,
      "loss": 5.9894,
      "mean_token_accuracy": 0.13505751192569732,
      "num_tokens": 4556877.0,
      "step": 2620
    },
    {
      "entropy": 6.094733190536499,
      "epoch": 0.20424042015172145,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998199430966481,
      "loss": 5.8389,
      "mean_token_accuracy": 0.1471896141767502,
      "num_tokens": 4565999.0,
      "step": 2625
    },
    {
      "entropy": 6.09189739227295,
      "epoch": 0.20462944952343903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998188328141346,
      "loss": 5.8909,
      "mean_token_accuracy": 0.13818741962313652,
      "num_tokens": 4574027.0,
      "step": 2630
    },
    {
      "entropy": 6.029260110855103,
      "epoch": 0.2050184788951566,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998177191203531,
      "loss": 5.8962,
      "mean_token_accuracy": 0.1333038367331028,
      "num_tokens": 4582402.0,
      "step": 2635
    },
    {
      "entropy": 6.102605962753296,
      "epoch": 0.20540750826687415,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998166020153204,
      "loss": 5.8411,
      "mean_token_accuracy": 0.14166108146309853,
      "num_tokens": 4590381.0,
      "step": 2640
    },
    {
      "entropy": 6.096075630187988,
      "epoch": 0.2057965376385917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998154814990538,
      "loss": 5.9002,
      "mean_token_accuracy": 0.13965460956096648,
      "num_tokens": 4598910.0,
      "step": 2645
    },
    {
      "entropy": 6.230145454406738,
      "epoch": 0.20618556701030927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049981435757157,
      "loss": 6.0215,
      "mean_token_accuracy": 0.13467366695404054,
      "num_tokens": 4607400.0,
      "step": 2650
    },
    {
      "entropy": 6.057853126525879,
      "epoch": 0.20657459638202685,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998132302328862,
      "loss": 5.9545,
      "mean_token_accuracy": 0.14162048250436782,
      "num_tokens": 4616510.0,
      "step": 2655
    },
    {
      "entropy": 6.152973604202271,
      "epoch": 0.2069636257537444,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998120994830195,
      "loss": 5.9042,
      "mean_token_accuracy": 0.14345652610063553,
      "num_tokens": 4624529.0,
      "step": 2660
    },
    {
      "entropy": 6.153080034255981,
      "epoch": 0.20735265512546197,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499810965321987,
      "loss": 5.8473,
      "mean_token_accuracy": 0.144081724435091,
      "num_tokens": 4632141.0,
      "step": 2665
    },
    {
      "entropy": 6.135512018203736,
      "epoch": 0.20774168449717953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998098277498059,
      "loss": 6.0399,
      "mean_token_accuracy": 0.13816287219524384,
      "num_tokens": 4640378.0,
      "step": 2670
    },
    {
      "entropy": 6.064649534225464,
      "epoch": 0.2081307138688971,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998086867664934,
      "loss": 5.8117,
      "mean_token_accuracy": 0.1414397269487381,
      "num_tokens": 4649313.0,
      "step": 2675
    },
    {
      "entropy": 6.121465873718262,
      "epoch": 0.20851974324061467,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499807542372067,
      "loss": 5.9071,
      "mean_token_accuracy": 0.14331126660108567,
      "num_tokens": 4658272.0,
      "step": 2680
    },
    {
      "entropy": 6.036688280105591,
      "epoch": 0.20890877261233223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998063945665439,
      "loss": 5.7901,
      "mean_token_accuracy": 0.1520828664302826,
      "num_tokens": 4665970.0,
      "step": 2685
    },
    {
      "entropy": 6.189170980453492,
      "epoch": 0.2092978019840498,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998052433499416,
      "loss": 6.0293,
      "mean_token_accuracy": 0.13662858083844184,
      "num_tokens": 4675308.0,
      "step": 2690
    },
    {
      "entropy": 6.077781295776367,
      "epoch": 0.20968683135576735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998040887222776,
      "loss": 5.8742,
      "mean_token_accuracy": 0.14358604997396468,
      "num_tokens": 4684316.0,
      "step": 2695
    },
    {
      "entropy": 6.031701469421387,
      "epoch": 0.21007586072748494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998029306835693,
      "loss": 5.7824,
      "mean_token_accuracy": 0.14697564840316774,
      "num_tokens": 4693029.0,
      "step": 2700
    },
    {
      "entropy": 6.0618767738342285,
      "epoch": 0.2104648900992025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998017692338344,
      "loss": 5.9842,
      "mean_token_accuracy": 0.13347845301032066,
      "num_tokens": 4702404.0,
      "step": 2705
    },
    {
      "entropy": 6.141680812835693,
      "epoch": 0.21085391947092005,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998006043730905,
      "loss": 5.9526,
      "mean_token_accuracy": 0.13373763784766196,
      "num_tokens": 4711126.0,
      "step": 2710
    },
    {
      "entropy": 6.116554880142212,
      "epoch": 0.2112429488426376,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997994361013551,
      "loss": 5.9516,
      "mean_token_accuracy": 0.1492009460926056,
      "num_tokens": 4719816.0,
      "step": 2715
    },
    {
      "entropy": 6.174014472961426,
      "epoch": 0.21163197821435517,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997982644186461,
      "loss": 5.8835,
      "mean_token_accuracy": 0.14039621651172637,
      "num_tokens": 4728294.0,
      "step": 2720
    },
    {
      "entropy": 6.0613030910491945,
      "epoch": 0.21202100758607276,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997970893249813,
      "loss": 5.815,
      "mean_token_accuracy": 0.1434742122888565,
      "num_tokens": 4736708.0,
      "step": 2725
    },
    {
      "entropy": 6.044667100906372,
      "epoch": 0.21241003695779032,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997959108203785,
      "loss": 5.9964,
      "mean_token_accuracy": 0.1296418972313404,
      "num_tokens": 4744893.0,
      "step": 2730
    },
    {
      "entropy": 6.243737983703613,
      "epoch": 0.21279906632950787,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997947289048554,
      "loss": 5.9463,
      "mean_token_accuracy": 0.1440265618264675,
      "num_tokens": 4753246.0,
      "step": 2735
    },
    {
      "entropy": 6.026007032394409,
      "epoch": 0.21318809570122543,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997935435784302,
      "loss": 5.875,
      "mean_token_accuracy": 0.14041776210069656,
      "num_tokens": 4762229.0,
      "step": 2740
    },
    {
      "entropy": 5.99229941368103,
      "epoch": 0.21357712507294302,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997923548411208,
      "loss": 5.8139,
      "mean_token_accuracy": 0.148726749420166,
      "num_tokens": 4771315.0,
      "step": 2745
    },
    {
      "entropy": 5.956744146347046,
      "epoch": 0.21396615444466058,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499791162692945,
      "loss": 5.7628,
      "mean_token_accuracy": 0.15489397719502448,
      "num_tokens": 4779848.0,
      "step": 2750
    },
    {
      "entropy": 6.160625457763672,
      "epoch": 0.21435518381637814,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997899671339214,
      "loss": 5.999,
      "mean_token_accuracy": 0.13992350101470946,
      "num_tokens": 4788939.0,
      "step": 2755
    },
    {
      "entropy": 6.131371355056762,
      "epoch": 0.2147442131880957,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997887681640675,
      "loss": 5.9017,
      "mean_token_accuracy": 0.14416032284498215,
      "num_tokens": 4797719.0,
      "step": 2760
    },
    {
      "entropy": 6.09780068397522,
      "epoch": 0.21513324255981325,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997875657834021,
      "loss": 5.9332,
      "mean_token_accuracy": 0.13998101651668549,
      "num_tokens": 4805601.0,
      "step": 2765
    },
    {
      "entropy": 6.137105464935303,
      "epoch": 0.21552227193153084,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997863599919432,
      "loss": 5.8913,
      "mean_token_accuracy": 0.1408081702888012,
      "num_tokens": 4814993.0,
      "step": 2770
    },
    {
      "entropy": 6.115447616577148,
      "epoch": 0.2159113013032484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997851507897089,
      "loss": 5.8477,
      "mean_token_accuracy": 0.1397731989622116,
      "num_tokens": 4824156.0,
      "step": 2775
    },
    {
      "entropy": 5.991670274734497,
      "epoch": 0.21630033067496596,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997839381767178,
      "loss": 5.9414,
      "mean_token_accuracy": 0.14050936251878737,
      "num_tokens": 4833827.0,
      "step": 2780
    },
    {
      "entropy": 6.081092882156372,
      "epoch": 0.21668936004668352,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997827221529882,
      "loss": 5.8829,
      "mean_token_accuracy": 0.1428877130150795,
      "num_tokens": 4842801.0,
      "step": 2785
    },
    {
      "entropy": 6.08880844116211,
      "epoch": 0.2170783894184011,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997815027185386,
      "loss": 5.7536,
      "mean_token_accuracy": 0.14653713181614875,
      "num_tokens": 4850928.0,
      "step": 2790
    },
    {
      "entropy": 5.995206022262574,
      "epoch": 0.21746741879011866,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997802798733874,
      "loss": 5.8637,
      "mean_token_accuracy": 0.1513894386589527,
      "num_tokens": 4859510.0,
      "step": 2795
    },
    {
      "entropy": 6.10595588684082,
      "epoch": 0.21785644816183622,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997790536175534,
      "loss": 5.8902,
      "mean_token_accuracy": 0.1465415373444557,
      "num_tokens": 4868201.0,
      "step": 2800
    },
    {
      "entropy": 6.035786819458008,
      "epoch": 0.21824547753355378,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997778239510548,
      "loss": 5.816,
      "mean_token_accuracy": 0.15248959213495256,
      "num_tokens": 4876896.0,
      "step": 2805
    },
    {
      "entropy": 6.001589441299439,
      "epoch": 0.21863450690527134,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997765908739106,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1534420743584633,
      "num_tokens": 4885136.0,
      "step": 2810
    },
    {
      "entropy": 6.0229329586029055,
      "epoch": 0.21902353627698892,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997753543861395,
      "loss": 5.7893,
      "mean_token_accuracy": 0.14993626177310942,
      "num_tokens": 4893277.0,
      "step": 2815
    },
    {
      "entropy": 6.036485242843628,
      "epoch": 0.21941256564870648,
      "grad_norm": 1.0,
      "learning_rate": 0.00049977411448776,
      "loss": 5.9514,
      "mean_token_accuracy": 0.1361225239932537,
      "num_tokens": 4901977.0,
      "step": 2820
    },
    {
      "entropy": 6.142391729354858,
      "epoch": 0.21980159502042404,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997728711787912,
      "loss": 5.9431,
      "mean_token_accuracy": 0.1336057133972645,
      "num_tokens": 4910888.0,
      "step": 2825
    },
    {
      "entropy": 6.031327724456787,
      "epoch": 0.2201906243921416,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997716244592518,
      "loss": 5.9456,
      "mean_token_accuracy": 0.13697238340973855,
      "num_tokens": 4920369.0,
      "step": 2830
    },
    {
      "entropy": 6.1665221691131595,
      "epoch": 0.22057965376385918,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997703743291607,
      "loss": 5.9347,
      "mean_token_accuracy": 0.13784253150224685,
      "num_tokens": 4928799.0,
      "step": 2835
    },
    {
      "entropy": 6.118892431259155,
      "epoch": 0.22096868313557674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499769120788537,
      "loss": 5.8422,
      "mean_token_accuracy": 0.15058550387620925,
      "num_tokens": 4937194.0,
      "step": 2840
    },
    {
      "entropy": 6.045169401168823,
      "epoch": 0.2213577125072943,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997678638373995,
      "loss": 5.9202,
      "mean_token_accuracy": 0.13822367861866952,
      "num_tokens": 4946174.0,
      "step": 2845
    },
    {
      "entropy": 6.097653436660766,
      "epoch": 0.22174674187901186,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997666034757676,
      "loss": 5.8543,
      "mean_token_accuracy": 0.13650616109371186,
      "num_tokens": 4955376.0,
      "step": 2850
    },
    {
      "entropy": 6.003319025039673,
      "epoch": 0.22213577125072942,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997653397036603,
      "loss": 5.8545,
      "mean_token_accuracy": 0.14963753297924995,
      "num_tokens": 4964138.0,
      "step": 2855
    },
    {
      "entropy": 6.087133550643921,
      "epoch": 0.222524800622447,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997640725210965,
      "loss": 5.945,
      "mean_token_accuracy": 0.13923729509115218,
      "num_tokens": 4973868.0,
      "step": 2860
    },
    {
      "entropy": 6.072530794143677,
      "epoch": 0.22291382999416456,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997628019280958,
      "loss": 5.8158,
      "mean_token_accuracy": 0.14615875333547593,
      "num_tokens": 4982687.0,
      "step": 2865
    },
    {
      "entropy": 6.07082724571228,
      "epoch": 0.22330285936588212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997615279246773,
      "loss": 5.8724,
      "mean_token_accuracy": 0.13691072314977645,
      "num_tokens": 4992072.0,
      "step": 2870
    },
    {
      "entropy": 6.080668687820435,
      "epoch": 0.22369188873759968,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997602505108603,
      "loss": 5.8261,
      "mean_token_accuracy": 0.14304942935705184,
      "num_tokens": 5000551.0,
      "step": 2875
    },
    {
      "entropy": 6.073841190338134,
      "epoch": 0.22408091810931724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997589696866644,
      "loss": 5.9859,
      "mean_token_accuracy": 0.13843553885817528,
      "num_tokens": 5009292.0,
      "step": 2880
    },
    {
      "entropy": 6.066045570373535,
      "epoch": 0.22446994748103483,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997576854521088,
      "loss": 5.8342,
      "mean_token_accuracy": 0.1451627016067505,
      "num_tokens": 5018407.0,
      "step": 2885
    },
    {
      "entropy": 5.999263858795166,
      "epoch": 0.22485897685275238,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499756397807213,
      "loss": 5.8877,
      "mean_token_accuracy": 0.14157619029283525,
      "num_tokens": 5027353.0,
      "step": 2890
    },
    {
      "entropy": 6.1293292999267575,
      "epoch": 0.22524800622446994,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997551067519966,
      "loss": 5.8898,
      "mean_token_accuracy": 0.14480420351028442,
      "num_tokens": 5036476.0,
      "step": 2895
    },
    {
      "entropy": 6.099696016311645,
      "epoch": 0.2256370355961875,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997538122864792,
      "loss": 5.8458,
      "mean_token_accuracy": 0.1427316889166832,
      "num_tokens": 5045132.0,
      "step": 2900
    },
    {
      "entropy": 6.082901287078857,
      "epoch": 0.2260260649679051,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997525144106804,
      "loss": 5.8717,
      "mean_token_accuracy": 0.1474373534321785,
      "num_tokens": 5053665.0,
      "step": 2905
    },
    {
      "entropy": 6.110235738754272,
      "epoch": 0.22641509433962265,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049975121312462,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1529403179883957,
      "num_tokens": 5062516.0,
      "step": 2910
    },
    {
      "entropy": 6.043049669265747,
      "epoch": 0.2268041237113402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997499084283177,
      "loss": 5.8536,
      "mean_token_accuracy": 0.14332823008298873,
      "num_tokens": 5071260.0,
      "step": 2915
    },
    {
      "entropy": 5.952953195571899,
      "epoch": 0.22719315308305776,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997486003217932,
      "loss": 5.8082,
      "mean_token_accuracy": 0.15106913149356843,
      "num_tokens": 5079215.0,
      "step": 2920
    },
    {
      "entropy": 6.071919727325439,
      "epoch": 0.22758218245477532,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997472888050664,
      "loss": 5.8879,
      "mean_token_accuracy": 0.1414925940334797,
      "num_tokens": 5088404.0,
      "step": 2925
    },
    {
      "entropy": 6.055942630767822,
      "epoch": 0.2279712118264929,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997459738781573,
      "loss": 5.803,
      "mean_token_accuracy": 0.147076266258955,
      "num_tokens": 5096756.0,
      "step": 2930
    },
    {
      "entropy": 6.042949390411377,
      "epoch": 0.22836024119821047,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997446555410857,
      "loss": 5.967,
      "mean_token_accuracy": 0.14154459834098815,
      "num_tokens": 5106160.0,
      "step": 2935
    },
    {
      "entropy": 5.963182592391968,
      "epoch": 0.22874927056992803,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997433337938716,
      "loss": 5.899,
      "mean_token_accuracy": 0.14055694192647933,
      "num_tokens": 5114915.0,
      "step": 2940
    },
    {
      "entropy": 6.1321821212768555,
      "epoch": 0.22913829994164558,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997420086365351,
      "loss": 5.8087,
      "mean_token_accuracy": 0.14655397981405258,
      "num_tokens": 5123249.0,
      "step": 2945
    },
    {
      "entropy": 5.994801664352417,
      "epoch": 0.22952732931336317,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997406800690966,
      "loss": 5.7989,
      "mean_token_accuracy": 0.14508022367954254,
      "num_tokens": 5131906.0,
      "step": 2950
    },
    {
      "entropy": 6.047790002822876,
      "epoch": 0.22991635868508073,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997393480915758,
      "loss": 5.795,
      "mean_token_accuracy": 0.1496501974761486,
      "num_tokens": 5140740.0,
      "step": 2955
    },
    {
      "entropy": 6.069603347778321,
      "epoch": 0.2303053880567983,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997380127039931,
      "loss": 5.9802,
      "mean_token_accuracy": 0.14554987102746964,
      "num_tokens": 5149748.0,
      "step": 2960
    },
    {
      "entropy": 5.9712036609649655,
      "epoch": 0.23069441742851585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997366739063688,
      "loss": 5.7259,
      "mean_token_accuracy": 0.15179628431797026,
      "num_tokens": 5158475.0,
      "step": 2965
    },
    {
      "entropy": 6.014142608642578,
      "epoch": 0.2310834468002334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997353316987231,
      "loss": 5.9024,
      "mean_token_accuracy": 0.1394285723567009,
      "num_tokens": 5166593.0,
      "step": 2970
    },
    {
      "entropy": 6.108148241043091,
      "epoch": 0.231472476171951,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997339860810765,
      "loss": 5.8795,
      "mean_token_accuracy": 0.1452465608716011,
      "num_tokens": 5175557.0,
      "step": 2975
    },
    {
      "entropy": 6.0830058574676515,
      "epoch": 0.23186150554366855,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997326370534495,
      "loss": 5.7837,
      "mean_token_accuracy": 0.14554769918322563,
      "num_tokens": 5184140.0,
      "step": 2980
    },
    {
      "entropy": 6.019090270996093,
      "epoch": 0.2322505349153861,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997312846158622,
      "loss": 5.8684,
      "mean_token_accuracy": 0.14349435344338418,
      "num_tokens": 5192347.0,
      "step": 2985
    },
    {
      "entropy": 6.08717942237854,
      "epoch": 0.23263956428710367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997299287683355,
      "loss": 5.8239,
      "mean_token_accuracy": 0.14950033277273178,
      "num_tokens": 5201327.0,
      "step": 2990
    },
    {
      "entropy": 6.0416545391082765,
      "epoch": 0.23302859365882125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997285695108898,
      "loss": 5.9213,
      "mean_token_accuracy": 0.14163968116044998,
      "num_tokens": 5209873.0,
      "step": 2995
    },
    {
      "entropy": 6.0891759395599365,
      "epoch": 0.2334176230305388,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997272068435458,
      "loss": 5.8338,
      "mean_token_accuracy": 0.14707664102315904,
      "num_tokens": 5219176.0,
      "step": 3000
    },
    {
      "epoch": 0.2334176230305388,
      "eval_entropy": 5.810826111115207,
      "eval_loss": 5.8746232986450195,
      "eval_mean_token_accuracy": 0.15242935873360103,
      "eval_num_tokens": 5219176.0,
      "eval_runtime": 21.8651,
      "eval_samples_per_second": 1900.651,
      "eval_steps_per_second": 237.593,
      "step": 3000
    },
    {
      "entropy": 5.952131080627441,
      "epoch": 0.23380665240225637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499725840766324,
      "loss": 5.7713,
      "mean_token_accuracy": 0.15496642887592316,
      "num_tokens": 5227310.0,
      "step": 3005
    },
    {
      "entropy": 6.058289813995361,
      "epoch": 0.23419568177397393,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997244712792453,
      "loss": 5.8553,
      "mean_token_accuracy": 0.14385973513126374,
      "num_tokens": 5235882.0,
      "step": 3010
    },
    {
      "entropy": 6.016433668136597,
      "epoch": 0.2345847111456915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997230983823305,
      "loss": 5.7775,
      "mean_token_accuracy": 0.1440333232283592,
      "num_tokens": 5244149.0,
      "step": 3015
    },
    {
      "entropy": 5.913664770126343,
      "epoch": 0.23497374051740907,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997217220756003,
      "loss": 5.6384,
      "mean_token_accuracy": 0.15045991390943528,
      "num_tokens": 5252473.0,
      "step": 3020
    },
    {
      "entropy": 5.9935009479522705,
      "epoch": 0.23536276988912663,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997203423590757,
      "loss": 5.7945,
      "mean_token_accuracy": 0.15354376584291457,
      "num_tokens": 5260972.0,
      "step": 3025
    },
    {
      "entropy": 5.997221040725708,
      "epoch": 0.2357517992608442,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997189592327775,
      "loss": 5.818,
      "mean_token_accuracy": 0.14062808826565742,
      "num_tokens": 5270513.0,
      "step": 3030
    },
    {
      "entropy": 6.049147224426269,
      "epoch": 0.23614082863256175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997175726967268,
      "loss": 5.7648,
      "mean_token_accuracy": 0.1412667602300644,
      "num_tokens": 5278433.0,
      "step": 3035
    },
    {
      "entropy": 5.951753997802735,
      "epoch": 0.23652985800427934,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997161827509447,
      "loss": 5.7927,
      "mean_token_accuracy": 0.14049935191869736,
      "num_tokens": 5287389.0,
      "step": 3040
    },
    {
      "entropy": 6.017498970031738,
      "epoch": 0.2369188873759969,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997147893954521,
      "loss": 5.9,
      "mean_token_accuracy": 0.1400333970785141,
      "num_tokens": 5296220.0,
      "step": 3045
    },
    {
      "entropy": 6.065495729446411,
      "epoch": 0.23730791674771445,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997133926302702,
      "loss": 5.8294,
      "mean_token_accuracy": 0.14628155678510665,
      "num_tokens": 5304152.0,
      "step": 3050
    },
    {
      "entropy": 6.058481073379516,
      "epoch": 0.237696946119432,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997119924554204,
      "loss": 5.8375,
      "mean_token_accuracy": 0.14466165602207184,
      "num_tokens": 5312374.0,
      "step": 3055
    },
    {
      "entropy": 5.962250900268555,
      "epoch": 0.23808597549114957,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997105888709236,
      "loss": 5.7717,
      "mean_token_accuracy": 0.1462335154414177,
      "num_tokens": 5320696.0,
      "step": 3060
    },
    {
      "entropy": 6.089827919006348,
      "epoch": 0.23847500486286716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997091818768015,
      "loss": 5.8991,
      "mean_token_accuracy": 0.14249396994709967,
      "num_tokens": 5328754.0,
      "step": 3065
    },
    {
      "entropy": 6.049135971069336,
      "epoch": 0.23886403423458472,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499707771473075,
      "loss": 5.8798,
      "mean_token_accuracy": 0.14481773376464843,
      "num_tokens": 5337553.0,
      "step": 3070
    },
    {
      "entropy": 6.059122467041016,
      "epoch": 0.23925306360630227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997063576597659,
      "loss": 5.8083,
      "mean_token_accuracy": 0.14893458187580108,
      "num_tokens": 5346063.0,
      "step": 3075
    },
    {
      "entropy": 6.010602426528931,
      "epoch": 0.23964209297801983,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997049404368954,
      "loss": 5.8362,
      "mean_token_accuracy": 0.14832401573657988,
      "num_tokens": 5354187.0,
      "step": 3080
    },
    {
      "entropy": 6.053317737579346,
      "epoch": 0.2400311223497374,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499703519804485,
      "loss": 5.8967,
      "mean_token_accuracy": 0.13576578721404076,
      "num_tokens": 5362819.0,
      "step": 3085
    },
    {
      "entropy": 6.0170165538787845,
      "epoch": 0.24042015172145498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997020957625564,
      "loss": 5.7775,
      "mean_token_accuracy": 0.1515388771891594,
      "num_tokens": 5371249.0,
      "step": 3090
    },
    {
      "entropy": 5.961355829238892,
      "epoch": 0.24080918109317254,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997006683111312,
      "loss": 5.7771,
      "mean_token_accuracy": 0.1478361189365387,
      "num_tokens": 5379284.0,
      "step": 3095
    },
    {
      "entropy": 6.0198503017425535,
      "epoch": 0.2411982104648901,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499699237450231,
      "loss": 5.8656,
      "mean_token_accuracy": 0.14309499114751817,
      "num_tokens": 5388541.0,
      "step": 3100
    },
    {
      "entropy": 6.061077070236206,
      "epoch": 0.24158723983660765,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996978031798774,
      "loss": 5.9518,
      "mean_token_accuracy": 0.14579177871346474,
      "num_tokens": 5397226.0,
      "step": 3105
    },
    {
      "entropy": 6.082793474197388,
      "epoch": 0.24197626920832524,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996963655000924,
      "loss": 5.8529,
      "mean_token_accuracy": 0.1392069309949875,
      "num_tokens": 5405579.0,
      "step": 3110
    },
    {
      "entropy": 6.0679725170135494,
      "epoch": 0.2423652985800428,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996949244108977,
      "loss": 5.9556,
      "mean_token_accuracy": 0.13417180627584457,
      "num_tokens": 5414529.0,
      "step": 3115
    },
    {
      "entropy": 6.133157062530517,
      "epoch": 0.24275432795176036,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499693479912315,
      "loss": 5.9106,
      "mean_token_accuracy": 0.1521434135735035,
      "num_tokens": 5423641.0,
      "step": 3120
    },
    {
      "entropy": 5.969631338119507,
      "epoch": 0.24314335732347792,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996920320043666,
      "loss": 5.8221,
      "mean_token_accuracy": 0.14464605897665023,
      "num_tokens": 5432558.0,
      "step": 3125
    },
    {
      "entropy": 6.00016598701477,
      "epoch": 0.24353238669519547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996905806870742,
      "loss": 5.761,
      "mean_token_accuracy": 0.14678752720355986,
      "num_tokens": 5441449.0,
      "step": 3130
    },
    {
      "entropy": 5.984752225875854,
      "epoch": 0.24392141606691306,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004996891259604598,
      "loss": 5.7856,
      "mean_token_accuracy": 0.14641717076301575,
      "num_tokens": 5451142.0,
      "step": 3135
    },
    {
      "entropy": 6.0575848579406735,
      "epoch": 0.24431044543863062,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996876678245455,
      "loss": 5.7395,
      "mean_token_accuracy": 0.1428086817264557,
      "num_tokens": 5459290.0,
      "step": 3140
    },
    {
      "entropy": 5.957022380828858,
      "epoch": 0.24469947481034818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996862062793536,
      "loss": 5.7646,
      "mean_token_accuracy": 0.14870759695768357,
      "num_tokens": 5467776.0,
      "step": 3145
    },
    {
      "entropy": 5.925604677200317,
      "epoch": 0.24508850418206574,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996847413249061,
      "loss": 5.6371,
      "mean_token_accuracy": 0.15342139899730683,
      "num_tokens": 5476005.0,
      "step": 3150
    },
    {
      "entropy": 6.05728669166565,
      "epoch": 0.24547753355378332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996832729612252,
      "loss": 5.8323,
      "mean_token_accuracy": 0.14415311068296432,
      "num_tokens": 5484225.0,
      "step": 3155
    },
    {
      "entropy": 5.874774217605591,
      "epoch": 0.24586656292550088,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996818011883333,
      "loss": 5.7544,
      "mean_token_accuracy": 0.15134995579719543,
      "num_tokens": 5493111.0,
      "step": 3160
    },
    {
      "entropy": 5.95450005531311,
      "epoch": 0.24625559229721844,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996803260062527,
      "loss": 5.7695,
      "mean_token_accuracy": 0.1509148046374321,
      "num_tokens": 5501574.0,
      "step": 3165
    },
    {
      "entropy": 6.052458000183106,
      "epoch": 0.246644621668936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996788474150057,
      "loss": 5.8094,
      "mean_token_accuracy": 0.14301597848534583,
      "num_tokens": 5511136.0,
      "step": 3170
    },
    {
      "entropy": 5.959417247772217,
      "epoch": 0.24703365104065356,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996773654146149,
      "loss": 5.7911,
      "mean_token_accuracy": 0.1535992830991745,
      "num_tokens": 5520324.0,
      "step": 3175
    },
    {
      "entropy": 6.005702638626099,
      "epoch": 0.24742268041237114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996758800051026,
      "loss": 5.7357,
      "mean_token_accuracy": 0.15172705948352813,
      "num_tokens": 5528960.0,
      "step": 3180
    },
    {
      "entropy": 5.977138042449951,
      "epoch": 0.2478117097840887,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004996743911864916,
      "loss": 5.8783,
      "mean_token_accuracy": 0.1430502340197563,
      "num_tokens": 5537391.0,
      "step": 3185
    },
    {
      "entropy": 6.041247272491455,
      "epoch": 0.24820073915580626,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996728989588041,
      "loss": 5.8273,
      "mean_token_accuracy": 0.14786960929632187,
      "num_tokens": 5546577.0,
      "step": 3190
    },
    {
      "entropy": 6.11059718132019,
      "epoch": 0.24858976852752382,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996714033220632,
      "loss": 5.7734,
      "mean_token_accuracy": 0.1480727419257164,
      "num_tokens": 5555247.0,
      "step": 3195
    },
    {
      "entropy": 5.937183904647827,
      "epoch": 0.2489787978992414,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996699042762911,
      "loss": 5.7373,
      "mean_token_accuracy": 0.1495134323835373,
      "num_tokens": 5563598.0,
      "step": 3200
    },
    {
      "entropy": 5.994940948486328,
      "epoch": 0.24936782727095896,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996684018215111,
      "loss": 5.7687,
      "mean_token_accuracy": 0.14412149786949158,
      "num_tokens": 5572734.0,
      "step": 3205
    },
    {
      "entropy": 6.040948915481567,
      "epoch": 0.24975685664267652,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996668959577453,
      "loss": 5.8818,
      "mean_token_accuracy": 0.13702815026044846,
      "num_tokens": 5581224.0,
      "step": 3210
    },
    {
      "entropy": 6.036828708648682,
      "epoch": 0.2501458860143941,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996653866850173,
      "loss": 5.874,
      "mean_token_accuracy": 0.14364129900932313,
      "num_tokens": 5590138.0,
      "step": 3215
    },
    {
      "entropy": 5.95453782081604,
      "epoch": 0.25053491538611167,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996638740033495,
      "loss": 5.804,
      "mean_token_accuracy": 0.15532443821430206,
      "num_tokens": 5598654.0,
      "step": 3220
    },
    {
      "entropy": 5.928944540023804,
      "epoch": 0.2509239447578292,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996623579127651,
      "loss": 5.7358,
      "mean_token_accuracy": 0.15299883037805556,
      "num_tokens": 5607404.0,
      "step": 3225
    },
    {
      "entropy": 6.012897586822509,
      "epoch": 0.2513129741295468,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996608384132868,
      "loss": 5.7861,
      "mean_token_accuracy": 0.1459243580698967,
      "num_tokens": 5617133.0,
      "step": 3230
    },
    {
      "entropy": 5.951272678375244,
      "epoch": 0.25170200350126437,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499659315504938,
      "loss": 5.8027,
      "mean_token_accuracy": 0.14748322814702988,
      "num_tokens": 5625389.0,
      "step": 3235
    },
    {
      "entropy": 5.944720315933227,
      "epoch": 0.2520910328729819,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996577891877417,
      "loss": 5.687,
      "mean_token_accuracy": 0.16281363517045974,
      "num_tokens": 5633985.0,
      "step": 3240
    },
    {
      "entropy": 5.99700517654419,
      "epoch": 0.2524800622446995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996562594617209,
      "loss": 5.796,
      "mean_token_accuracy": 0.14644584953784942,
      "num_tokens": 5642524.0,
      "step": 3245
    },
    {
      "entropy": 5.970114612579346,
      "epoch": 0.252869091616417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996547263268991,
      "loss": 5.7617,
      "mean_token_accuracy": 0.1493512749671936,
      "num_tokens": 5651788.0,
      "step": 3250
    },
    {
      "entropy": 5.963803434371949,
      "epoch": 0.2532581209881346,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996531897832993,
      "loss": 5.7994,
      "mean_token_accuracy": 0.14588565677404403,
      "num_tokens": 5660757.0,
      "step": 3255
    },
    {
      "entropy": 6.069337368011475,
      "epoch": 0.2536471503598522,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499651649830945,
      "loss": 5.8857,
      "mean_token_accuracy": 0.14849150702357292,
      "num_tokens": 5669760.0,
      "step": 3260
    },
    {
      "entropy": 5.976683855056763,
      "epoch": 0.2540361797315697,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996501064698593,
      "loss": 5.7987,
      "mean_token_accuracy": 0.14993155896663665,
      "num_tokens": 5678313.0,
      "step": 3265
    },
    {
      "entropy": 6.002733898162842,
      "epoch": 0.2544252091032873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499648559700066,
      "loss": 5.7386,
      "mean_token_accuracy": 0.1515405535697937,
      "num_tokens": 5686942.0,
      "step": 3270
    },
    {
      "entropy": 5.939556074142456,
      "epoch": 0.25481423847500484,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996470095215884,
      "loss": 5.7344,
      "mean_token_accuracy": 0.15350284799933434,
      "num_tokens": 5695612.0,
      "step": 3275
    },
    {
      "entropy": 5.9842628002166744,
      "epoch": 0.2552032678467224,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996454559344498,
      "loss": 5.9101,
      "mean_token_accuracy": 0.1402068942785263,
      "num_tokens": 5705067.0,
      "step": 3280
    },
    {
      "entropy": 6.016226005554199,
      "epoch": 0.25559229721844,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996438989386741,
      "loss": 5.7969,
      "mean_token_accuracy": 0.1526660218834877,
      "num_tokens": 5714092.0,
      "step": 3285
    },
    {
      "entropy": 6.013244676589966,
      "epoch": 0.25598132659015754,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996423385342847,
      "loss": 5.76,
      "mean_token_accuracy": 0.14619042873382568,
      "num_tokens": 5722638.0,
      "step": 3290
    },
    {
      "entropy": 5.914634847640992,
      "epoch": 0.25637035596187513,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996407747213055,
      "loss": 5.7679,
      "mean_token_accuracy": 0.1486264154314995,
      "num_tokens": 5731511.0,
      "step": 3295
    },
    {
      "entropy": 5.97312822341919,
      "epoch": 0.25675938533359266,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049963920749976,
      "loss": 5.7893,
      "mean_token_accuracy": 0.14694643169641494,
      "num_tokens": 5740950.0,
      "step": 3300
    },
    {
      "entropy": 5.965716123580933,
      "epoch": 0.25714841470531025,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996376368696721,
      "loss": 5.7409,
      "mean_token_accuracy": 0.14827728495001793,
      "num_tokens": 5749163.0,
      "step": 3305
    },
    {
      "entropy": 5.9895940780639645,
      "epoch": 0.25753744407702783,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996360628310656,
      "loss": 5.7993,
      "mean_token_accuracy": 0.14380623251199723,
      "num_tokens": 5758321.0,
      "step": 3310
    },
    {
      "entropy": 5.987315702438354,
      "epoch": 0.25792647344874536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996344853839644,
      "loss": 5.8318,
      "mean_token_accuracy": 0.15102650225162506,
      "num_tokens": 5767277.0,
      "step": 3315
    },
    {
      "entropy": 5.956531429290772,
      "epoch": 0.25831550282046295,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996329045283924,
      "loss": 5.7378,
      "mean_token_accuracy": 0.15711987465620042,
      "num_tokens": 5775694.0,
      "step": 3320
    },
    {
      "entropy": 5.862797784805298,
      "epoch": 0.25870453219218054,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996313202643737,
      "loss": 5.6754,
      "mean_token_accuracy": 0.15443939119577407,
      "num_tokens": 5783932.0,
      "step": 3325
    },
    {
      "entropy": 5.931440162658691,
      "epoch": 0.25909356156389807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996297325919321,
      "loss": 5.7765,
      "mean_token_accuracy": 0.15169637501239777,
      "num_tokens": 5792100.0,
      "step": 3330
    },
    {
      "entropy": 6.000880861282349,
      "epoch": 0.25948259093561565,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996281415110919,
      "loss": 5.7575,
      "mean_token_accuracy": 0.14606721848249435,
      "num_tokens": 5801481.0,
      "step": 3335
    },
    {
      "entropy": 6.011248159408569,
      "epoch": 0.2598716203073332,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996265470218773,
      "loss": 5.8665,
      "mean_token_accuracy": 0.13883595392107964,
      "num_tokens": 5810053.0,
      "step": 3340
    },
    {
      "entropy": 5.949085235595703,
      "epoch": 0.26026064967905077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996249491243122,
      "loss": 5.644,
      "mean_token_accuracy": 0.15924251973628997,
      "num_tokens": 5818454.0,
      "step": 3345
    },
    {
      "entropy": 5.955591249465942,
      "epoch": 0.26064967905076836,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996233478184211,
      "loss": 5.8055,
      "mean_token_accuracy": 0.1437770664691925,
      "num_tokens": 5827013.0,
      "step": 3350
    },
    {
      "entropy": 5.916650199890137,
      "epoch": 0.2610387084224859,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996217431042282,
      "loss": 5.8127,
      "mean_token_accuracy": 0.14451763406395912,
      "num_tokens": 5835862.0,
      "step": 3355
    },
    {
      "entropy": 5.91140513420105,
      "epoch": 0.2614277377942035,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499620134981758,
      "loss": 5.5945,
      "mean_token_accuracy": 0.15513415932655333,
      "num_tokens": 5843827.0,
      "step": 3360
    },
    {
      "entropy": 5.905231285095215,
      "epoch": 0.261816767165921,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996185234510346,
      "loss": 5.7856,
      "mean_token_accuracy": 0.14372429698705674,
      "num_tokens": 5852677.0,
      "step": 3365
    },
    {
      "entropy": 6.019457149505615,
      "epoch": 0.2622057965376386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996169085120828,
      "loss": 5.7293,
      "mean_token_accuracy": 0.1516408622264862,
      "num_tokens": 5861686.0,
      "step": 3370
    },
    {
      "entropy": 5.822810935974121,
      "epoch": 0.2625948259093562,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996152901649268,
      "loss": 5.7035,
      "mean_token_accuracy": 0.146698522567749,
      "num_tokens": 5870690.0,
      "step": 3375
    },
    {
      "entropy": 5.91604962348938,
      "epoch": 0.2629838552810737,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996136684095913,
      "loss": 5.7925,
      "mean_token_accuracy": 0.14702642261981963,
      "num_tokens": 5879804.0,
      "step": 3380
    },
    {
      "entropy": 6.05110535621643,
      "epoch": 0.2633728846527913,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996120432461009,
      "loss": 5.8651,
      "mean_token_accuracy": 0.14215986281633378,
      "num_tokens": 5888781.0,
      "step": 3385
    },
    {
      "entropy": 6.032693004608154,
      "epoch": 0.2637619140245088,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996104146744804,
      "loss": 5.7781,
      "mean_token_accuracy": 0.14683165550231933,
      "num_tokens": 5897439.0,
      "step": 3390
    },
    {
      "entropy": 5.953556299209595,
      "epoch": 0.2641509433962264,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996087826947541,
      "loss": 5.7296,
      "mean_token_accuracy": 0.15696240961551666,
      "num_tokens": 5905471.0,
      "step": 3395
    },
    {
      "entropy": 5.875892972946167,
      "epoch": 0.264539972767944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996071473069471,
      "loss": 5.7758,
      "mean_token_accuracy": 0.1474592626094818,
      "num_tokens": 5913795.0,
      "step": 3400
    },
    {
      "entropy": 6.075421142578125,
      "epoch": 0.26492900213966153,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996055085110842,
      "loss": 5.8241,
      "mean_token_accuracy": 0.14667866080999375,
      "num_tokens": 5922119.0,
      "step": 3405
    },
    {
      "entropy": 5.87319130897522,
      "epoch": 0.2653180315113791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996038663071902,
      "loss": 5.7083,
      "mean_token_accuracy": 0.15487584322690964,
      "num_tokens": 5930759.0,
      "step": 3410
    },
    {
      "entropy": 5.9472472190856935,
      "epoch": 0.26570706088309665,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996022206952901,
      "loss": 5.756,
      "mean_token_accuracy": 0.15228534936904908,
      "num_tokens": 5939830.0,
      "step": 3415
    },
    {
      "entropy": 6.013852500915528,
      "epoch": 0.26609609025481423,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996005716754086,
      "loss": 5.7532,
      "mean_token_accuracy": 0.1481878861784935,
      "num_tokens": 5947737.0,
      "step": 3420
    },
    {
      "entropy": 5.824540424346924,
      "epoch": 0.2664851196265318,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499598919247571,
      "loss": 5.7127,
      "mean_token_accuracy": 0.15554220527410506,
      "num_tokens": 5955905.0,
      "step": 3425
    },
    {
      "entropy": 5.882647848129272,
      "epoch": 0.26687414899824935,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995972634118023,
      "loss": 5.7799,
      "mean_token_accuracy": 0.14817242622375487,
      "num_tokens": 5965223.0,
      "step": 3430
    },
    {
      "entropy": 6.018713331222534,
      "epoch": 0.26726317836996694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995956041681277,
      "loss": 5.7758,
      "mean_token_accuracy": 0.1541240818798542,
      "num_tokens": 5973615.0,
      "step": 3435
    },
    {
      "entropy": 5.928654193878174,
      "epoch": 0.2676522077416845,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995939415165722,
      "loss": 5.7046,
      "mean_token_accuracy": 0.15365282595157623,
      "num_tokens": 5981600.0,
      "step": 3440
    },
    {
      "entropy": 5.9306252002716064,
      "epoch": 0.26804123711340205,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995922754571611,
      "loss": 5.7102,
      "mean_token_accuracy": 0.14901533499360084,
      "num_tokens": 5989928.0,
      "step": 3445
    },
    {
      "entropy": 5.887343263626098,
      "epoch": 0.26843026648511964,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995906059899197,
      "loss": 5.7794,
      "mean_token_accuracy": 0.148210871219635,
      "num_tokens": 5998704.0,
      "step": 3450
    },
    {
      "entropy": 5.988488817214966,
      "epoch": 0.26881929585683717,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995889331148733,
      "loss": 5.75,
      "mean_token_accuracy": 0.1488671362400055,
      "num_tokens": 6007476.0,
      "step": 3455
    },
    {
      "entropy": 5.996799659729004,
      "epoch": 0.26920832522855476,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995872568320474,
      "loss": 5.8222,
      "mean_token_accuracy": 0.14480241239070893,
      "num_tokens": 6016324.0,
      "step": 3460
    },
    {
      "entropy": 5.961569309234619,
      "epoch": 0.26959735460027234,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995855771414673,
      "loss": 5.7541,
      "mean_token_accuracy": 0.14858102351427077,
      "num_tokens": 6024709.0,
      "step": 3465
    },
    {
      "entropy": 5.877323961257934,
      "epoch": 0.2699863839719899,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995838940431585,
      "loss": 5.6257,
      "mean_token_accuracy": 0.15648279190063477,
      "num_tokens": 6033131.0,
      "step": 3470
    },
    {
      "entropy": 5.879639291763306,
      "epoch": 0.27037541334370746,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995822075371466,
      "loss": 5.8089,
      "mean_token_accuracy": 0.14873216822743415,
      "num_tokens": 6042310.0,
      "step": 3475
    },
    {
      "entropy": 5.997434377670288,
      "epoch": 0.270764442715425,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995805176234571,
      "loss": 5.6343,
      "mean_token_accuracy": 0.1559881567955017,
      "num_tokens": 6050617.0,
      "step": 3480
    },
    {
      "entropy": 5.933330535888672,
      "epoch": 0.2711534720871426,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995788243021158,
      "loss": 5.7966,
      "mean_token_accuracy": 0.1464315079152584,
      "num_tokens": 6060090.0,
      "step": 3485
    },
    {
      "entropy": 5.924555921554566,
      "epoch": 0.27154250145886016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995771275731483,
      "loss": 5.7479,
      "mean_token_accuracy": 0.15804181396961212,
      "num_tokens": 6068646.0,
      "step": 3490
    },
    {
      "entropy": 5.989955282211303,
      "epoch": 0.2719315308305777,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995754274365802,
      "loss": 5.7974,
      "mean_token_accuracy": 0.15477645993232728,
      "num_tokens": 6077896.0,
      "step": 3495
    },
    {
      "entropy": 5.899862575531006,
      "epoch": 0.2723205602022953,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995737238924376,
      "loss": 5.726,
      "mean_token_accuracy": 0.15056513994932175,
      "num_tokens": 6087189.0,
      "step": 3500
    },
    {
      "entropy": 5.922992134094239,
      "epoch": 0.2727095895740128,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995720169407461,
      "loss": 5.7228,
      "mean_token_accuracy": 0.15020885169506074,
      "num_tokens": 6095394.0,
      "step": 3505
    },
    {
      "entropy": 5.974159383773804,
      "epoch": 0.2730986189457304,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995703065815317,
      "loss": 5.6944,
      "mean_token_accuracy": 0.15587686151266097,
      "num_tokens": 6104135.0,
      "step": 3510
    },
    {
      "entropy": 5.9481529712677,
      "epoch": 0.273487648317448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995685928148203,
      "loss": 5.8156,
      "mean_token_accuracy": 0.14244434088468552,
      "num_tokens": 6112160.0,
      "step": 3515
    },
    {
      "entropy": 5.964101696014405,
      "epoch": 0.2738766776891655,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499566875640638,
      "loss": 5.7994,
      "mean_token_accuracy": 0.15270021110773085,
      "num_tokens": 6120575.0,
      "step": 3520
    },
    {
      "entropy": 5.9738319396972654,
      "epoch": 0.2742657070608831,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995651550590108,
      "loss": 5.8112,
      "mean_token_accuracy": 0.14792770594358445,
      "num_tokens": 6129367.0,
      "step": 3525
    },
    {
      "entropy": 5.962098455429077,
      "epoch": 0.2746547364326007,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995634310699647,
      "loss": 5.7942,
      "mean_token_accuracy": 0.15110692679882048,
      "num_tokens": 6137440.0,
      "step": 3530
    },
    {
      "entropy": 5.932923603057861,
      "epoch": 0.2750437658043182,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995617036735261,
      "loss": 5.7075,
      "mean_token_accuracy": 0.1477666363120079,
      "num_tokens": 6146770.0,
      "step": 3535
    },
    {
      "entropy": 6.0504333019256595,
      "epoch": 0.2754327951760358,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995599728697211,
      "loss": 5.7933,
      "mean_token_accuracy": 0.14865502268075942,
      "num_tokens": 6155994.0,
      "step": 3540
    },
    {
      "entropy": 5.864378118515015,
      "epoch": 0.27582182454775334,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995582386585759,
      "loss": 5.7062,
      "mean_token_accuracy": 0.14862505793571473,
      "num_tokens": 6163941.0,
      "step": 3545
    },
    {
      "entropy": 5.870920419692993,
      "epoch": 0.2762108539194709,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995565010401167,
      "loss": 5.7646,
      "mean_token_accuracy": 0.14493151903152465,
      "num_tokens": 6172681.0,
      "step": 3550
    },
    {
      "entropy": 5.996222305297851,
      "epoch": 0.2765998832911885,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995547600143702,
      "loss": 5.7357,
      "mean_token_accuracy": 0.14585817828774453,
      "num_tokens": 6181449.0,
      "step": 3555
    },
    {
      "entropy": 5.958917427062988,
      "epoch": 0.27698891266290604,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995530155813625,
      "loss": 5.7392,
      "mean_token_accuracy": 0.151982182264328,
      "num_tokens": 6190275.0,
      "step": 3560
    },
    {
      "entropy": 5.987959623336792,
      "epoch": 0.2773779420346236,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995512677411203,
      "loss": 5.7593,
      "mean_token_accuracy": 0.15145091116428375,
      "num_tokens": 6198521.0,
      "step": 3565
    },
    {
      "entropy": 5.9518767356872555,
      "epoch": 0.27776697140634116,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995495164936698,
      "loss": 5.8045,
      "mean_token_accuracy": 0.145390148460865,
      "num_tokens": 6207318.0,
      "step": 3570
    },
    {
      "entropy": 5.938201665878296,
      "epoch": 0.27815600077805874,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995477618390381,
      "loss": 5.6857,
      "mean_token_accuracy": 0.15625451654195785,
      "num_tokens": 6216693.0,
      "step": 3575
    },
    {
      "entropy": 5.929543781280517,
      "epoch": 0.27854503014977633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995460037772513,
      "loss": 5.7442,
      "mean_token_accuracy": 0.14824536591768264,
      "num_tokens": 6225207.0,
      "step": 3580
    },
    {
      "entropy": 5.921706628799439,
      "epoch": 0.27893405952149386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995442423083365,
      "loss": 5.6769,
      "mean_token_accuracy": 0.16117929220199584,
      "num_tokens": 6232653.0,
      "step": 3585
    },
    {
      "entropy": 5.865441989898682,
      "epoch": 0.27932308889321145,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995424774323201,
      "loss": 5.679,
      "mean_token_accuracy": 0.15469851717352867,
      "num_tokens": 6241093.0,
      "step": 3590
    },
    {
      "entropy": 5.866435813903808,
      "epoch": 0.279712118264929,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499540709149229,
      "loss": 5.6016,
      "mean_token_accuracy": 0.160512213408947,
      "num_tokens": 6249144.0,
      "step": 3595
    },
    {
      "entropy": 5.914186668395996,
      "epoch": 0.28010114763664656,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995389374590901,
      "loss": 5.6837,
      "mean_token_accuracy": 0.1514856621623039,
      "num_tokens": 6257622.0,
      "step": 3600
    },
    {
      "entropy": 5.9322789192199705,
      "epoch": 0.28049017700836415,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995371623619301,
      "loss": 5.702,
      "mean_token_accuracy": 0.15351740270853043,
      "num_tokens": 6266435.0,
      "step": 3605
    },
    {
      "entropy": 5.906361675262451,
      "epoch": 0.2808792063800817,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499535383857776,
      "loss": 5.8196,
      "mean_token_accuracy": 0.14777727127075196,
      "num_tokens": 6275053.0,
      "step": 3610
    },
    {
      "entropy": 5.903389263153076,
      "epoch": 0.28126823575179927,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499533601946655,
      "loss": 5.7206,
      "mean_token_accuracy": 0.14442056640982628,
      "num_tokens": 6283240.0,
      "step": 3615
    },
    {
      "entropy": 5.860544919967651,
      "epoch": 0.2816572651235168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995318166285938,
      "loss": 5.6591,
      "mean_token_accuracy": 0.15395397394895555,
      "num_tokens": 6292133.0,
      "step": 3620
    },
    {
      "entropy": 6.009118270874024,
      "epoch": 0.2820462944952344,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995300279036199,
      "loss": 5.7813,
      "mean_token_accuracy": 0.1485302358865738,
      "num_tokens": 6300954.0,
      "step": 3625
    },
    {
      "entropy": 5.8872617244720455,
      "epoch": 0.28243532386695197,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049952823577176,
      "loss": 5.6629,
      "mean_token_accuracy": 0.15599143356084824,
      "num_tokens": 6309637.0,
      "step": 3630
    },
    {
      "entropy": 5.881065750122071,
      "epoch": 0.2828243532386695,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995264402330416,
      "loss": 5.731,
      "mean_token_accuracy": 0.15445464551448823,
      "num_tokens": 6318539.0,
      "step": 3635
    },
    {
      "entropy": 5.94295105934143,
      "epoch": 0.2832133826103871,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995246412874919,
      "loss": 5.6383,
      "mean_token_accuracy": 0.1552157461643219,
      "num_tokens": 6327109.0,
      "step": 3640
    },
    {
      "entropy": 5.852169609069824,
      "epoch": 0.2836024119821047,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499522838935138,
      "loss": 5.6472,
      "mean_token_accuracy": 0.15341811031103134,
      "num_tokens": 6335555.0,
      "step": 3645
    },
    {
      "entropy": 5.9907721996307375,
      "epoch": 0.2839914413538222,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995210331760074,
      "loss": 5.8083,
      "mean_token_accuracy": 0.14825723096728324,
      "num_tokens": 6344716.0,
      "step": 3650
    },
    {
      "entropy": 5.878650665283203,
      "epoch": 0.2843804707255398,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995192240101275,
      "loss": 5.6731,
      "mean_token_accuracy": 0.15794774815440177,
      "num_tokens": 6353618.0,
      "step": 3655
    },
    {
      "entropy": 5.955765151977539,
      "epoch": 0.2847695000972573,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995174114375258,
      "loss": 5.7819,
      "mean_token_accuracy": 0.14444806203246116,
      "num_tokens": 6362039.0,
      "step": 3660
    },
    {
      "entropy": 5.974342250823975,
      "epoch": 0.2851585294689749,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995155954582297,
      "loss": 5.7043,
      "mean_token_accuracy": 0.15271616280078887,
      "num_tokens": 6370331.0,
      "step": 3665
    },
    {
      "entropy": 5.874191999435425,
      "epoch": 0.2855475588406925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995137760722668,
      "loss": 5.6893,
      "mean_token_accuracy": 0.1580304831266403,
      "num_tokens": 6378943.0,
      "step": 3670
    },
    {
      "entropy": 5.931831502914429,
      "epoch": 0.28593658821241,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995119532796646,
      "loss": 5.7531,
      "mean_token_accuracy": 0.14501482248306274,
      "num_tokens": 6387745.0,
      "step": 3675
    },
    {
      "entropy": 5.767093324661255,
      "epoch": 0.2863256175841276,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499510127080451,
      "loss": 5.5946,
      "mean_token_accuracy": 0.15369964838027955,
      "num_tokens": 6396915.0,
      "step": 3680
    },
    {
      "entropy": 5.857335329055786,
      "epoch": 0.28671464695584514,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995082974746535,
      "loss": 5.6422,
      "mean_token_accuracy": 0.15733037143945694,
      "num_tokens": 6405453.0,
      "step": 3685
    },
    {
      "entropy": 5.993175363540649,
      "epoch": 0.28710367632756273,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995064644622999,
      "loss": 5.7256,
      "mean_token_accuracy": 0.15438124239444734,
      "num_tokens": 6413693.0,
      "step": 3690
    },
    {
      "entropy": 5.838237667083741,
      "epoch": 0.2874927056992803,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995046280434181,
      "loss": 5.7059,
      "mean_token_accuracy": 0.1526482358574867,
      "num_tokens": 6422738.0,
      "step": 3695
    },
    {
      "entropy": 5.890118265151978,
      "epoch": 0.28788173507099785,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995027882180358,
      "loss": 5.6564,
      "mean_token_accuracy": 0.15297402441501617,
      "num_tokens": 6431743.0,
      "step": 3700
    },
    {
      "entropy": 5.965820693969727,
      "epoch": 0.28827076444271543,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995009449861812,
      "loss": 5.72,
      "mean_token_accuracy": 0.14820689782500268,
      "num_tokens": 6440586.0,
      "step": 3705
    },
    {
      "entropy": 5.840203714370728,
      "epoch": 0.28865979381443296,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499499098347882,
      "loss": 5.7455,
      "mean_token_accuracy": 0.15329755693674088,
      "num_tokens": 6449644.0,
      "step": 3710
    },
    {
      "entropy": 5.87632384300232,
      "epoch": 0.28904882318615055,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994972483031662,
      "loss": 5.6352,
      "mean_token_accuracy": 0.15695199966430665,
      "num_tokens": 6458752.0,
      "step": 3715
    },
    {
      "entropy": 5.858027935028076,
      "epoch": 0.28943785255786814,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499495394852062,
      "loss": 5.7095,
      "mean_token_accuracy": 0.15834027528762817,
      "num_tokens": 6468333.0,
      "step": 3720
    },
    {
      "entropy": 5.7514075756073,
      "epoch": 0.28982688192958567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994935379945977,
      "loss": 5.5545,
      "mean_token_accuracy": 0.1531543478369713,
      "num_tokens": 6477406.0,
      "step": 3725
    },
    {
      "entropy": 5.944380474090576,
      "epoch": 0.29021591130130325,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994916777308012,
      "loss": 5.7195,
      "mean_token_accuracy": 0.150412417948246,
      "num_tokens": 6486399.0,
      "step": 3730
    },
    {
      "entropy": 5.872902822494507,
      "epoch": 0.29060494067302084,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994898140607007,
      "loss": 5.6269,
      "mean_token_accuracy": 0.15273130536079407,
      "num_tokens": 6494853.0,
      "step": 3735
    },
    {
      "entropy": 5.8865861892700195,
      "epoch": 0.29099397004473837,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994879469843247,
      "loss": 5.7928,
      "mean_token_accuracy": 0.15120636373758317,
      "num_tokens": 6503301.0,
      "step": 3740
    },
    {
      "entropy": 5.914371299743652,
      "epoch": 0.29138299941645596,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994860765017013,
      "loss": 5.655,
      "mean_token_accuracy": 0.15607353746891023,
      "num_tokens": 6512685.0,
      "step": 3745
    },
    {
      "entropy": 5.927185344696045,
      "epoch": 0.2917720287881735,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499484202612859,
      "loss": 5.7769,
      "mean_token_accuracy": 0.14915118962526322,
      "num_tokens": 6520673.0,
      "step": 3750
    },
    {
      "entropy": 5.938401174545288,
      "epoch": 0.2921610581598911,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994823253178264,
      "loss": 5.6896,
      "mean_token_accuracy": 0.14857900589704515,
      "num_tokens": 6529039.0,
      "step": 3755
    },
    {
      "entropy": 5.914093446731568,
      "epoch": 0.29255008753160866,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994804446166317,
      "loss": 5.7737,
      "mean_token_accuracy": 0.14460716545581817,
      "num_tokens": 6537932.0,
      "step": 3760
    },
    {
      "entropy": 5.911954402923584,
      "epoch": 0.2929391169033262,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994785605093035,
      "loss": 5.6603,
      "mean_token_accuracy": 0.15702853798866273,
      "num_tokens": 6546278.0,
      "step": 3765
    },
    {
      "entropy": 5.901885318756103,
      "epoch": 0.2933281462750438,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994766729958705,
      "loss": 5.7168,
      "mean_token_accuracy": 0.1562728241086006,
      "num_tokens": 6554449.0,
      "step": 3770
    },
    {
      "entropy": 5.899599218368531,
      "epoch": 0.2937171756467613,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994747820763612,
      "loss": 5.6971,
      "mean_token_accuracy": 0.1506945237517357,
      "num_tokens": 6563272.0,
      "step": 3775
    },
    {
      "entropy": 5.957968664169312,
      "epoch": 0.2941062050184789,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994728877508046,
      "loss": 5.7129,
      "mean_token_accuracy": 0.15193629264831543,
      "num_tokens": 6571864.0,
      "step": 3780
    },
    {
      "entropy": 5.884029769897461,
      "epoch": 0.2944952343901965,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994709900192289,
      "loss": 5.7189,
      "mean_token_accuracy": 0.1543201759457588,
      "num_tokens": 6580435.0,
      "step": 3785
    },
    {
      "entropy": 5.945118761062622,
      "epoch": 0.294884263761914,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994690888816635,
      "loss": 5.6991,
      "mean_token_accuracy": 0.15705762356519698,
      "num_tokens": 6589261.0,
      "step": 3790
    },
    {
      "entropy": 5.926768493652344,
      "epoch": 0.2952732931336316,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994671843381367,
      "loss": 5.7893,
      "mean_token_accuracy": 0.15674624741077423,
      "num_tokens": 6598057.0,
      "step": 3795
    },
    {
      "entropy": 6.038666200637818,
      "epoch": 0.29566232250534913,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994652763886778,
      "loss": 5.7407,
      "mean_token_accuracy": 0.14874251037836075,
      "num_tokens": 6606439.0,
      "step": 3800
    },
    {
      "entropy": 5.909378576278686,
      "epoch": 0.2960513518770667,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994633650333155,
      "loss": 5.759,
      "mean_token_accuracy": 0.15097366273403168,
      "num_tokens": 6615350.0,
      "step": 3805
    },
    {
      "entropy": 5.94745659828186,
      "epoch": 0.2964403812487843,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994614502720792,
      "loss": 5.7999,
      "mean_token_accuracy": 0.15037863925099373,
      "num_tokens": 6624258.0,
      "step": 3810
    },
    {
      "entropy": 6.008837270736694,
      "epoch": 0.29682941062050183,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994595321049974,
      "loss": 5.8022,
      "mean_token_accuracy": 0.14758660942316054,
      "num_tokens": 6632382.0,
      "step": 3815
    },
    {
      "entropy": 5.907459926605225,
      "epoch": 0.2972184399922194,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994576105320994,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1514850899577141,
      "num_tokens": 6640482.0,
      "step": 3820
    },
    {
      "entropy": 5.891694784164429,
      "epoch": 0.29760746936393695,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994556855534146,
      "loss": 5.6966,
      "mean_token_accuracy": 0.1510847233235836,
      "num_tokens": 6648912.0,
      "step": 3825
    },
    {
      "entropy": 5.872078132629395,
      "epoch": 0.29799649873565454,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994537571689719,
      "loss": 5.6701,
      "mean_token_accuracy": 0.15535251945257186,
      "num_tokens": 6656912.0,
      "step": 3830
    },
    {
      "entropy": 5.794346189498901,
      "epoch": 0.2983855281073721,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994518253788007,
      "loss": 5.6881,
      "mean_token_accuracy": 0.15521788746118545,
      "num_tokens": 6665900.0,
      "step": 3835
    },
    {
      "entropy": 5.958291530609131,
      "epoch": 0.29877455747908965,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994498901829304,
      "loss": 5.7269,
      "mean_token_accuracy": 0.1525902658700943,
      "num_tokens": 6674426.0,
      "step": 3840
    },
    {
      "entropy": 5.920195150375366,
      "epoch": 0.29916358685080724,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994479515813903,
      "loss": 5.695,
      "mean_token_accuracy": 0.1515626683831215,
      "num_tokens": 6683566.0,
      "step": 3845
    },
    {
      "entropy": 5.899086236953735,
      "epoch": 0.2995526162225248,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994460095742096,
      "loss": 5.706,
      "mean_token_accuracy": 0.15467460155487062,
      "num_tokens": 6693341.0,
      "step": 3850
    },
    {
      "entropy": 5.868103265762329,
      "epoch": 0.29994164559424236,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499444064161418,
      "loss": 5.6215,
      "mean_token_accuracy": 0.15215489268302917,
      "num_tokens": 6701734.0,
      "step": 3855
    },
    {
      "entropy": 5.824473190307617,
      "epoch": 0.30033067496595994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499442115343045,
      "loss": 5.6365,
      "mean_token_accuracy": 0.1546481318771839,
      "num_tokens": 6710572.0,
      "step": 3860
    },
    {
      "entropy": 5.957624721527099,
      "epoch": 0.3007197043376775,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994401631191202,
      "loss": 5.7714,
      "mean_token_accuracy": 0.15227345675230025,
      "num_tokens": 6719355.0,
      "step": 3865
    },
    {
      "entropy": 6.054756021499633,
      "epoch": 0.30110873370939506,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994382074896731,
      "loss": 5.8059,
      "mean_token_accuracy": 0.1459599368274212,
      "num_tokens": 6729279.0,
      "step": 3870
    },
    {
      "entropy": 5.869380950927734,
      "epoch": 0.30149776308111265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994362484547335,
      "loss": 5.6276,
      "mean_token_accuracy": 0.15225827246904372,
      "num_tokens": 6737318.0,
      "step": 3875
    },
    {
      "entropy": 5.836999082565308,
      "epoch": 0.3018867924528302,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994342860143309,
      "loss": 5.697,
      "mean_token_accuracy": 0.15468793660402297,
      "num_tokens": 6746520.0,
      "step": 3880
    },
    {
      "entropy": 5.906482934951782,
      "epoch": 0.30227582182454776,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994323201684953,
      "loss": 5.6324,
      "mean_token_accuracy": 0.15672309696674347,
      "num_tokens": 6755685.0,
      "step": 3885
    },
    {
      "entropy": 5.908680152893067,
      "epoch": 0.3026648511962653,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994303509172566,
      "loss": 5.7777,
      "mean_token_accuracy": 0.15273956432938576,
      "num_tokens": 6764347.0,
      "step": 3890
    },
    {
      "entropy": 5.960914659500122,
      "epoch": 0.3030538805679829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994283782606444,
      "loss": 5.7472,
      "mean_token_accuracy": 0.15492507815361023,
      "num_tokens": 6772887.0,
      "step": 3895
    },
    {
      "entropy": 5.8742760181427,
      "epoch": 0.30344290993970047,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994264021986888,
      "loss": 5.7094,
      "mean_token_accuracy": 0.15225957781076432,
      "num_tokens": 6780860.0,
      "step": 3900
    },
    {
      "entropy": 5.863096094131469,
      "epoch": 0.303831939311418,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994244227314197,
      "loss": 5.7264,
      "mean_token_accuracy": 0.1514242857694626,
      "num_tokens": 6789491.0,
      "step": 3905
    },
    {
      "entropy": 5.935235166549683,
      "epoch": 0.3042209686831356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994224398588672,
      "loss": 5.6535,
      "mean_token_accuracy": 0.1513040542602539,
      "num_tokens": 6797128.0,
      "step": 3910
    },
    {
      "entropy": 5.891047668457031,
      "epoch": 0.3046099980548531,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994204535810615,
      "loss": 5.6909,
      "mean_token_accuracy": 0.15431376546621323,
      "num_tokens": 6805235.0,
      "step": 3915
    },
    {
      "entropy": 5.868199157714844,
      "epoch": 0.3049990274265707,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994184638980326,
      "loss": 5.6084,
      "mean_token_accuracy": 0.1601799726486206,
      "num_tokens": 6813863.0,
      "step": 3920
    },
    {
      "entropy": 5.863962411880493,
      "epoch": 0.3053880567982883,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994164708098107,
      "loss": 5.6776,
      "mean_token_accuracy": 0.15048581063747407,
      "num_tokens": 6822015.0,
      "step": 3925
    },
    {
      "entropy": 5.939310455322266,
      "epoch": 0.3057770861700058,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499414474316426,
      "loss": 5.7358,
      "mean_token_accuracy": 0.15054322332143782,
      "num_tokens": 6830598.0,
      "step": 3930
    },
    {
      "entropy": 5.945909214019776,
      "epoch": 0.3061661155417234,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994124744179088,
      "loss": 5.7162,
      "mean_token_accuracy": 0.15882268771529198,
      "num_tokens": 6839129.0,
      "step": 3935
    },
    {
      "entropy": 5.8954814910888675,
      "epoch": 0.306555144913441,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994104711142896,
      "loss": 5.6922,
      "mean_token_accuracy": 0.14843768924474715,
      "num_tokens": 6848199.0,
      "step": 3940
    },
    {
      "entropy": 5.881297302246094,
      "epoch": 0.3069441742851585,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994084644055986,
      "loss": 5.6994,
      "mean_token_accuracy": 0.15382420867681504,
      "num_tokens": 6857340.0,
      "step": 3945
    },
    {
      "entropy": 5.874563407897949,
      "epoch": 0.3073332036568761,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994064542918664,
      "loss": 5.7174,
      "mean_token_accuracy": 0.14996173679828645,
      "num_tokens": 6865830.0,
      "step": 3950
    },
    {
      "entropy": 5.900748682022095,
      "epoch": 0.30772223302859364,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994044407731235,
      "loss": 5.7746,
      "mean_token_accuracy": 0.14810442328453063,
      "num_tokens": 6874318.0,
      "step": 3955
    },
    {
      "entropy": 6.012032747268677,
      "epoch": 0.3081112624003112,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994024238494002,
      "loss": 5.7755,
      "mean_token_accuracy": 0.14986879080533982,
      "num_tokens": 6882008.0,
      "step": 3960
    },
    {
      "entropy": 5.889507865905761,
      "epoch": 0.3085002917720288,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994004035207274,
      "loss": 5.6253,
      "mean_token_accuracy": 0.1561218023300171,
      "num_tokens": 6890411.0,
      "step": 3965
    },
    {
      "entropy": 5.893429088592529,
      "epoch": 0.30888932114374634,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993983797871356,
      "loss": 5.6826,
      "mean_token_accuracy": 0.1497599869966507,
      "num_tokens": 6899201.0,
      "step": 3970
    },
    {
      "entropy": 5.840681505203247,
      "epoch": 0.30927835051546393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993963526486555,
      "loss": 5.6858,
      "mean_token_accuracy": 0.15194668620824814,
      "num_tokens": 6907470.0,
      "step": 3975
    },
    {
      "entropy": 5.852532529830933,
      "epoch": 0.30966737988718146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499394322105318,
      "loss": 5.7041,
      "mean_token_accuracy": 0.14796101227402686,
      "num_tokens": 6916542.0,
      "step": 3980
    },
    {
      "entropy": 5.925456285476685,
      "epoch": 0.31005640925889905,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993922881571538,
      "loss": 5.7059,
      "mean_token_accuracy": 0.1543884590268135,
      "num_tokens": 6925974.0,
      "step": 3985
    },
    {
      "entropy": 5.940940046310425,
      "epoch": 0.31044543863061663,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993902508041939,
      "loss": 5.7214,
      "mean_token_accuracy": 0.1485102429986,
      "num_tokens": 6934987.0,
      "step": 3990
    },
    {
      "entropy": 5.833662796020508,
      "epoch": 0.31083446800233416,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499388210046469,
      "loss": 5.616,
      "mean_token_accuracy": 0.15158363580703735,
      "num_tokens": 6944497.0,
      "step": 3995
    },
    {
      "entropy": 5.947600507736206,
      "epoch": 0.31122349737405175,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993861658840102,
      "loss": 5.7561,
      "mean_token_accuracy": 0.14682230204343796,
      "num_tokens": 6953394.0,
      "step": 4000
    },
    {
      "entropy": 5.91348032951355,
      "epoch": 0.3116125267457693,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993841183168484,
      "loss": 5.6792,
      "mean_token_accuracy": 0.15258519500494003,
      "num_tokens": 6961715.0,
      "step": 4005
    },
    {
      "entropy": 5.816651439666748,
      "epoch": 0.31200155611748687,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993820673450148,
      "loss": 5.6746,
      "mean_token_accuracy": 0.15342024117708206,
      "num_tokens": 6970041.0,
      "step": 4010
    },
    {
      "entropy": 5.982955694198608,
      "epoch": 0.31239058548920445,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993800129685404,
      "loss": 5.8009,
      "mean_token_accuracy": 0.14771148636937143,
      "num_tokens": 6979522.0,
      "step": 4015
    },
    {
      "entropy": 5.913341808319092,
      "epoch": 0.312779614860922,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993779551874565,
      "loss": 5.7701,
      "mean_token_accuracy": 0.14592012614011765,
      "num_tokens": 6988470.0,
      "step": 4020
    },
    {
      "entropy": 5.980531883239746,
      "epoch": 0.31316864423263957,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993758940017943,
      "loss": 5.7829,
      "mean_token_accuracy": 0.15340054780244827,
      "num_tokens": 6996696.0,
      "step": 4025
    },
    {
      "entropy": 5.934103012084961,
      "epoch": 0.31355767360435716,
      "grad_norm": 1.0,
      "learning_rate": 0.000499373829411585,
      "loss": 5.6555,
      "mean_token_accuracy": 0.1566230520606041,
      "num_tokens": 7005297.0,
      "step": 4030
    },
    {
      "entropy": 5.836995553970337,
      "epoch": 0.3139467029760747,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049937176141686,
      "loss": 5.6571,
      "mean_token_accuracy": 0.15415221452713013,
      "num_tokens": 7013569.0,
      "step": 4035
    },
    {
      "entropy": 5.85465898513794,
      "epoch": 0.3143357323477923,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993696900176506,
      "loss": 5.5989,
      "mean_token_accuracy": 0.1616859517991543,
      "num_tokens": 7021449.0,
      "step": 4040
    },
    {
      "entropy": 5.913087320327759,
      "epoch": 0.3147247617195098,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993676152139883,
      "loss": 5.723,
      "mean_token_accuracy": 0.15687182694673538,
      "num_tokens": 7029663.0,
      "step": 4045
    },
    {
      "entropy": 5.899198532104492,
      "epoch": 0.3151137910912274,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993655370059046,
      "loss": 5.7236,
      "mean_token_accuracy": 0.1520445629954338,
      "num_tokens": 7039038.0,
      "step": 4050
    },
    {
      "entropy": 5.860358858108521,
      "epoch": 0.315502820462945,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993634553934309,
      "loss": 5.6804,
      "mean_token_accuracy": 0.1576730951666832,
      "num_tokens": 7047595.0,
      "step": 4055
    },
    {
      "entropy": 5.895573568344116,
      "epoch": 0.3158918498346625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993613703765989,
      "loss": 5.6944,
      "mean_token_accuracy": 0.14608515426516533,
      "num_tokens": 7055711.0,
      "step": 4060
    },
    {
      "entropy": 5.915502405166626,
      "epoch": 0.3162808792063801,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993592819554402,
      "loss": 5.6557,
      "mean_token_accuracy": 0.15424971133470536,
      "num_tokens": 7064100.0,
      "step": 4065
    },
    {
      "entropy": 5.878922367095948,
      "epoch": 0.3166699085780976,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993571901299866,
      "loss": 5.6262,
      "mean_token_accuracy": 0.15816935151815414,
      "num_tokens": 7073040.0,
      "step": 4070
    },
    {
      "entropy": 5.89038724899292,
      "epoch": 0.3170589379498152,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993550949002697,
      "loss": 5.732,
      "mean_token_accuracy": 0.1527765415608883,
      "num_tokens": 7081513.0,
      "step": 4075
    },
    {
      "entropy": 5.812891912460327,
      "epoch": 0.3174479673215328,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993529962663213,
      "loss": 5.6021,
      "mean_token_accuracy": 0.1610344111919403,
      "num_tokens": 7090704.0,
      "step": 4080
    },
    {
      "entropy": 5.856521511077881,
      "epoch": 0.31783699669325033,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993508942281732,
      "loss": 5.7864,
      "mean_token_accuracy": 0.14460856541991235,
      "num_tokens": 7099692.0,
      "step": 4085
    },
    {
      "entropy": 5.857594585418701,
      "epoch": 0.3182260260649679,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993487887858575,
      "loss": 5.6008,
      "mean_token_accuracy": 0.15564313605427743,
      "num_tokens": 7108021.0,
      "step": 4090
    },
    {
      "entropy": 5.86187219619751,
      "epoch": 0.31861505543668545,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499346679939406,
      "loss": 5.6836,
      "mean_token_accuracy": 0.15484025478363037,
      "num_tokens": 7116689.0,
      "step": 4095
    },
    {
      "entropy": 5.89621319770813,
      "epoch": 0.31900408480840303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993445676888507,
      "loss": 5.7549,
      "mean_token_accuracy": 0.15203065276145936,
      "num_tokens": 7125453.0,
      "step": 4100
    },
    {
      "entropy": 6.004023694992066,
      "epoch": 0.3193931141801206,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993424520342236,
      "loss": 5.8134,
      "mean_token_accuracy": 0.14793906584382058,
      "num_tokens": 7133692.0,
      "step": 4105
    },
    {
      "entropy": 5.879203033447266,
      "epoch": 0.31978214355183815,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993403329755569,
      "loss": 5.6284,
      "mean_token_accuracy": 0.15951790809631347,
      "num_tokens": 7142362.0,
      "step": 4110
    },
    {
      "entropy": 5.824657297134399,
      "epoch": 0.32017117292355574,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993382105128828,
      "loss": 5.6397,
      "mean_token_accuracy": 0.1585059329867363,
      "num_tokens": 7151020.0,
      "step": 4115
    },
    {
      "entropy": 5.8254087448120115,
      "epoch": 0.32056020229527327,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993360846462333,
      "loss": 5.5615,
      "mean_token_accuracy": 0.1588086098432541,
      "num_tokens": 7159453.0,
      "step": 4120
    },
    {
      "entropy": 5.922162771224976,
      "epoch": 0.32094923166699085,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993339553756408,
      "loss": 5.8176,
      "mean_token_accuracy": 0.14753687456250192,
      "num_tokens": 7168758.0,
      "step": 4125
    },
    {
      "entropy": 5.877033710479736,
      "epoch": 0.32133826103870844,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993318227011378,
      "loss": 5.6291,
      "mean_token_accuracy": 0.1572786748409271,
      "num_tokens": 7177835.0,
      "step": 4130
    },
    {
      "entropy": 5.892328500747681,
      "epoch": 0.32172729041042597,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993296866227562,
      "loss": 5.6901,
      "mean_token_accuracy": 0.15761512517929077,
      "num_tokens": 7186424.0,
      "step": 4135
    },
    {
      "entropy": 5.856100797653198,
      "epoch": 0.32211631978214356,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993275471405288,
      "loss": 5.6535,
      "mean_token_accuracy": 0.16344444453716278,
      "num_tokens": 7194800.0,
      "step": 4140
    },
    {
      "entropy": 5.783264923095703,
      "epoch": 0.32250534915386114,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993254042544879,
      "loss": 5.5827,
      "mean_token_accuracy": 0.1560390807688236,
      "num_tokens": 7203820.0,
      "step": 4145
    },
    {
      "entropy": 5.820183610916137,
      "epoch": 0.3228943785255787,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499323257964666,
      "loss": 5.5752,
      "mean_token_accuracy": 0.15613881796598433,
      "num_tokens": 7211793.0,
      "step": 4150
    },
    {
      "entropy": 5.845516014099121,
      "epoch": 0.32328340789729626,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993211082710957,
      "loss": 5.64,
      "mean_token_accuracy": 0.1555764816701412,
      "num_tokens": 7220471.0,
      "step": 4155
    },
    {
      "entropy": 5.845905923843384,
      "epoch": 0.3236724372690138,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993189551738097,
      "loss": 5.6999,
      "mean_token_accuracy": 0.15712527334690093,
      "num_tokens": 7228788.0,
      "step": 4160
    },
    {
      "entropy": 5.932910299301147,
      "epoch": 0.3240614666407314,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993167986728405,
      "loss": 5.8487,
      "mean_token_accuracy": 0.14985144287347793,
      "num_tokens": 7237707.0,
      "step": 4165
    },
    {
      "entropy": 5.967710685729981,
      "epoch": 0.32445049601244896,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499314638768221,
      "loss": 5.6047,
      "mean_token_accuracy": 0.15651564300060272,
      "num_tokens": 7246296.0,
      "step": 4170
    },
    {
      "entropy": 5.778722095489502,
      "epoch": 0.3248395253841665,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993124754599839,
      "loss": 5.5444,
      "mean_token_accuracy": 0.1577083557844162,
      "num_tokens": 7255076.0,
      "step": 4175
    },
    {
      "entropy": 5.741897439956665,
      "epoch": 0.3252285547558841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993103087481619,
      "loss": 5.6365,
      "mean_token_accuracy": 0.15567630231380464,
      "num_tokens": 7263169.0,
      "step": 4180
    },
    {
      "entropy": 5.91665210723877,
      "epoch": 0.3256175841276016,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993081386327881,
      "loss": 5.6936,
      "mean_token_accuracy": 0.1606354683637619,
      "num_tokens": 7271748.0,
      "step": 4185
    },
    {
      "entropy": 5.820067644119263,
      "epoch": 0.3260066134993192,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993059651138953,
      "loss": 5.5103,
      "mean_token_accuracy": 0.15937861055135727,
      "num_tokens": 7280291.0,
      "step": 4190
    },
    {
      "entropy": 5.80422739982605,
      "epoch": 0.3263956428710368,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993037881915165,
      "loss": 5.6926,
      "mean_token_accuracy": 0.1541864365339279,
      "num_tokens": 7290307.0,
      "step": 4195
    },
    {
      "entropy": 5.916877174377442,
      "epoch": 0.3267846722427543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993016078656847,
      "loss": 5.794,
      "mean_token_accuracy": 0.14723934307694436,
      "num_tokens": 7297942.0,
      "step": 4200
    },
    {
      "entropy": 5.803368663787841,
      "epoch": 0.3271737016144719,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499299424136433,
      "loss": 5.6603,
      "mean_token_accuracy": 0.1538266658782959,
      "num_tokens": 7307378.0,
      "step": 4205
    },
    {
      "entropy": 5.864370203018188,
      "epoch": 0.32756273098618943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992972370037946,
      "loss": 5.6186,
      "mean_token_accuracy": 0.16045687794685365,
      "num_tokens": 7315985.0,
      "step": 4210
    },
    {
      "entropy": 5.85097484588623,
      "epoch": 0.327951760357907,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992950464678026,
      "loss": 5.733,
      "mean_token_accuracy": 0.15445947051048278,
      "num_tokens": 7324142.0,
      "step": 4215
    },
    {
      "entropy": 5.816914701461792,
      "epoch": 0.3283407897296246,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992928525284903,
      "loss": 5.595,
      "mean_token_accuracy": 0.15655487030744553,
      "num_tokens": 7333012.0,
      "step": 4220
    },
    {
      "entropy": 5.8839775085449215,
      "epoch": 0.32872981910134214,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499290655185891,
      "loss": 5.6154,
      "mean_token_accuracy": 0.15756229162216187,
      "num_tokens": 7341838.0,
      "step": 4225
    },
    {
      "entropy": 5.82137017250061,
      "epoch": 0.3291188484730597,
      "grad_norm": 1.0,
      "learning_rate": 0.000499288454440038,
      "loss": 5.5691,
      "mean_token_accuracy": 0.16859470158815384,
      "num_tokens": 7349986.0,
      "step": 4230
    },
    {
      "entropy": 5.914486122131348,
      "epoch": 0.3295078778447773,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992862502909645,
      "loss": 5.8116,
      "mean_token_accuracy": 0.14698196202516556,
      "num_tokens": 7359428.0,
      "step": 4235
    },
    {
      "entropy": 5.927544593811035,
      "epoch": 0.32989690721649484,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992840427387043,
      "loss": 5.7162,
      "mean_token_accuracy": 0.15304372012615203,
      "num_tokens": 7368524.0,
      "step": 4240
    },
    {
      "entropy": 5.8727219581604,
      "epoch": 0.3302859365882124,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992818317832908,
      "loss": 5.6085,
      "mean_token_accuracy": 0.15719811767339706,
      "num_tokens": 7377979.0,
      "step": 4245
    },
    {
      "entropy": 5.768430566787719,
      "epoch": 0.33067496595992996,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992796174247574,
      "loss": 5.5275,
      "mean_token_accuracy": 0.1579056575894356,
      "num_tokens": 7386283.0,
      "step": 4250
    },
    {
      "entropy": 5.815568542480468,
      "epoch": 0.33106399533164754,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992773996631378,
      "loss": 5.6545,
      "mean_token_accuracy": 0.14839731231331826,
      "num_tokens": 7395409.0,
      "step": 4255
    },
    {
      "entropy": 5.864266967773437,
      "epoch": 0.33145302470336513,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992751784984656,
      "loss": 5.727,
      "mean_token_accuracy": 0.15300896167755126,
      "num_tokens": 7403389.0,
      "step": 4260
    },
    {
      "entropy": 5.835431003570557,
      "epoch": 0.33184205407508266,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992729539307745,
      "loss": 5.6457,
      "mean_token_accuracy": 0.15788860768079757,
      "num_tokens": 7411506.0,
      "step": 4265
    },
    {
      "entropy": 5.941154909133911,
      "epoch": 0.33223108344680025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992707259600984,
      "loss": 5.7379,
      "mean_token_accuracy": 0.15247647762298583,
      "num_tokens": 7420219.0,
      "step": 4270
    },
    {
      "entropy": 5.904551935195923,
      "epoch": 0.3326201128185178,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992684945864709,
      "loss": 5.6486,
      "mean_token_accuracy": 0.15825132355093957,
      "num_tokens": 7428442.0,
      "step": 4275
    },
    {
      "entropy": 5.789218282699585,
      "epoch": 0.33300914219023536,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992662598099261,
      "loss": 5.6143,
      "mean_token_accuracy": 0.15656258910894394,
      "num_tokens": 7436364.0,
      "step": 4280
    },
    {
      "entropy": 5.89788179397583,
      "epoch": 0.33339817156195295,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992640216304975,
      "loss": 5.6913,
      "mean_token_accuracy": 0.14621787145733833,
      "num_tokens": 7444122.0,
      "step": 4285
    },
    {
      "entropy": 5.914680242538452,
      "epoch": 0.3337872009336705,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992617800482196,
      "loss": 5.7017,
      "mean_token_accuracy": 0.15293780714273453,
      "num_tokens": 7454122.0,
      "step": 4290
    },
    {
      "entropy": 5.853878164291382,
      "epoch": 0.33417623030538807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499259535063126,
      "loss": 5.6857,
      "mean_token_accuracy": 0.15544915497303008,
      "num_tokens": 7463057.0,
      "step": 4295
    },
    {
      "entropy": 5.905990076065064,
      "epoch": 0.3345652596771056,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499257286675251,
      "loss": 5.6639,
      "mean_token_accuracy": 0.15292353183031082,
      "num_tokens": 7471675.0,
      "step": 4300
    },
    {
      "entropy": 5.757207775115967,
      "epoch": 0.3349542890488232,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992550348846285,
      "loss": 5.6051,
      "mean_token_accuracy": 0.15792447626590728,
      "num_tokens": 7479870.0,
      "step": 4305
    },
    {
      "entropy": 5.861905193328857,
      "epoch": 0.33534331842054077,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992527796912928,
      "loss": 5.6753,
      "mean_token_accuracy": 0.1546726182103157,
      "num_tokens": 7489830.0,
      "step": 4310
    },
    {
      "entropy": 5.825132846832275,
      "epoch": 0.3357323477922583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992505210952782,
      "loss": 5.6057,
      "mean_token_accuracy": 0.16312549710273744,
      "num_tokens": 7497852.0,
      "step": 4315
    },
    {
      "entropy": 5.854568910598755,
      "epoch": 0.3361213771639759,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992482590966188,
      "loss": 5.7381,
      "mean_token_accuracy": 0.15417585670948028,
      "num_tokens": 7506419.0,
      "step": 4320
    },
    {
      "entropy": 5.936369276046753,
      "epoch": 0.3365104065356934,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992459936953489,
      "loss": 5.636,
      "mean_token_accuracy": 0.1566232457756996,
      "num_tokens": 7514757.0,
      "step": 4325
    },
    {
      "entropy": 5.812022876739502,
      "epoch": 0.336899435907411,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992437248915032,
      "loss": 5.689,
      "mean_token_accuracy": 0.1529418110847473,
      "num_tokens": 7523912.0,
      "step": 4330
    },
    {
      "entropy": 5.8119433403015135,
      "epoch": 0.3372884652791286,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992414526851157,
      "loss": 5.5805,
      "mean_token_accuracy": 0.1630261406302452,
      "num_tokens": 7532401.0,
      "step": 4335
    },
    {
      "entropy": 5.77750072479248,
      "epoch": 0.3376774946508461,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992391770762212,
      "loss": 5.5576,
      "mean_token_accuracy": 0.1565028890967369,
      "num_tokens": 7541257.0,
      "step": 4340
    },
    {
      "entropy": 5.755213880538941,
      "epoch": 0.3380665240225637,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499236898064854,
      "loss": 5.497,
      "mean_token_accuracy": 0.15672704726457595,
      "num_tokens": 7549812.0,
      "step": 4345
    },
    {
      "entropy": 5.798407077789307,
      "epoch": 0.3384555533942813,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992346156510488,
      "loss": 5.5914,
      "mean_token_accuracy": 0.15490633994340897,
      "num_tokens": 7559319.0,
      "step": 4350
    },
    {
      "entropy": 5.855447244644165,
      "epoch": 0.3388445827659988,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992323298348403,
      "loss": 5.7269,
      "mean_token_accuracy": 0.15112665742635728,
      "num_tokens": 7567900.0,
      "step": 4355
    },
    {
      "entropy": 5.900017166137696,
      "epoch": 0.3392336121377164,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499230040616263,
      "loss": 5.6568,
      "mean_token_accuracy": 0.14715979248285294,
      "num_tokens": 7576163.0,
      "step": 4360
    },
    {
      "entropy": 5.876814508438111,
      "epoch": 0.33962264150943394,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992277479953518,
      "loss": 5.5877,
      "mean_token_accuracy": 0.15960148274898528,
      "num_tokens": 7584221.0,
      "step": 4365
    },
    {
      "entropy": 5.785816764831543,
      "epoch": 0.34001167088115153,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992254519721414,
      "loss": 5.5811,
      "mean_token_accuracy": 0.15680823475122452,
      "num_tokens": 7593569.0,
      "step": 4370
    },
    {
      "entropy": 5.792740154266357,
      "epoch": 0.3404007002528691,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992231525466666,
      "loss": 5.6293,
      "mean_token_accuracy": 0.16478054970502853,
      "num_tokens": 7602666.0,
      "step": 4375
    },
    {
      "entropy": 5.784790182113648,
      "epoch": 0.34078972962458665,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992208497189624,
      "loss": 5.7135,
      "mean_token_accuracy": 0.14881878942251206,
      "num_tokens": 7611874.0,
      "step": 4380
    },
    {
      "entropy": 5.884900522232056,
      "epoch": 0.34117875899630423,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992185434890637,
      "loss": 5.6635,
      "mean_token_accuracy": 0.15799686759710313,
      "num_tokens": 7620385.0,
      "step": 4385
    },
    {
      "entropy": 5.953466272354126,
      "epoch": 0.34156778836802176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992162338570055,
      "loss": 5.7347,
      "mean_token_accuracy": 0.14705561548471452,
      "num_tokens": 7628786.0,
      "step": 4390
    },
    {
      "entropy": 5.810998582839966,
      "epoch": 0.34195681773973935,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992139208228227,
      "loss": 5.5345,
      "mean_token_accuracy": 0.1632216215133667,
      "num_tokens": 7637389.0,
      "step": 4395
    },
    {
      "entropy": 5.797764158248901,
      "epoch": 0.34234584711145694,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992116043865506,
      "loss": 5.6949,
      "mean_token_accuracy": 0.15499998927116393,
      "num_tokens": 7645793.0,
      "step": 4400
    },
    {
      "entropy": 5.909337186813355,
      "epoch": 0.34273487648317447,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992092845482244,
      "loss": 5.6644,
      "mean_token_accuracy": 0.15942255705595015,
      "num_tokens": 7653951.0,
      "step": 4405
    },
    {
      "entropy": 5.780727243423462,
      "epoch": 0.34312390585489205,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992069613078791,
      "loss": 5.5223,
      "mean_token_accuracy": 0.16487507224082948,
      "num_tokens": 7662753.0,
      "step": 4410
    },
    {
      "entropy": 5.742848300933838,
      "epoch": 0.3435129352266096,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992046346655499,
      "loss": 5.497,
      "mean_token_accuracy": 0.1634562462568283,
      "num_tokens": 7671832.0,
      "step": 4415
    },
    {
      "entropy": 5.833199977874756,
      "epoch": 0.34390196459832717,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992023046212723,
      "loss": 5.6674,
      "mean_token_accuracy": 0.15599291175603866,
      "num_tokens": 7680969.0,
      "step": 4420
    },
    {
      "entropy": 5.934517192840576,
      "epoch": 0.34429099397004476,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991999711750816,
      "loss": 5.6766,
      "mean_token_accuracy": 0.14898543506860734,
      "num_tokens": 7689277.0,
      "step": 4425
    },
    {
      "entropy": 5.823058843612671,
      "epoch": 0.3446800233417623,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991976343270133,
      "loss": 5.5828,
      "mean_token_accuracy": 0.1608385667204857,
      "num_tokens": 7697373.0,
      "step": 4430
    },
    {
      "entropy": 5.895063209533691,
      "epoch": 0.3450690527134799,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991952940771026,
      "loss": 5.6698,
      "mean_token_accuracy": 0.1545317880809307,
      "num_tokens": 7705767.0,
      "step": 4435
    },
    {
      "entropy": 5.798029994964599,
      "epoch": 0.34545808208519746,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991929504253852,
      "loss": 5.5428,
      "mean_token_accuracy": 0.1633567437529564,
      "num_tokens": 7713746.0,
      "step": 4440
    },
    {
      "entropy": 5.768852567672729,
      "epoch": 0.345847111456915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991906033718966,
      "loss": 5.6822,
      "mean_token_accuracy": 0.15390456914901735,
      "num_tokens": 7721829.0,
      "step": 4445
    },
    {
      "entropy": 5.790914487838745,
      "epoch": 0.3462361408286326,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991882529166724,
      "loss": 5.5554,
      "mean_token_accuracy": 0.1579376205801964,
      "num_tokens": 7730479.0,
      "step": 4450
    },
    {
      "entropy": 5.844191217422486,
      "epoch": 0.3466251702003501,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991858990597483,
      "loss": 5.5505,
      "mean_token_accuracy": 0.15994303971529006,
      "num_tokens": 7738659.0,
      "step": 4455
    },
    {
      "entropy": 5.898314809799194,
      "epoch": 0.3470141995720677,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991835418011601,
      "loss": 5.8052,
      "mean_token_accuracy": 0.15245812833309175,
      "num_tokens": 7747597.0,
      "step": 4460
    },
    {
      "entropy": 5.805671262741089,
      "epoch": 0.3474032289437853,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991811811409434,
      "loss": 5.6489,
      "mean_token_accuracy": 0.1574895977973938,
      "num_tokens": 7756666.0,
      "step": 4465
    },
    {
      "entropy": 5.818331956863403,
      "epoch": 0.3477922583155028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991788170791341,
      "loss": 5.6806,
      "mean_token_accuracy": 0.15582320243120193,
      "num_tokens": 7765621.0,
      "step": 4470
    },
    {
      "entropy": 5.80223822593689,
      "epoch": 0.3481812876872204,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991764496157682,
      "loss": 5.5815,
      "mean_token_accuracy": 0.1537637397646904,
      "num_tokens": 7774137.0,
      "step": 4475
    },
    {
      "entropy": 5.869473838806153,
      "epoch": 0.34857031705893793,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991740787508814,
      "loss": 5.6803,
      "mean_token_accuracy": 0.14955005794763565,
      "num_tokens": 7783144.0,
      "step": 4480
    },
    {
      "entropy": 5.82867841720581,
      "epoch": 0.3489593464306555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991717044845097,
      "loss": 5.7151,
      "mean_token_accuracy": 0.14956270828843116,
      "num_tokens": 7791530.0,
      "step": 4485
    },
    {
      "entropy": 5.81160798072815,
      "epoch": 0.3493483758023731,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991693268166892,
      "loss": 5.5698,
      "mean_token_accuracy": 0.1626602053642273,
      "num_tokens": 7800869.0,
      "step": 4490
    },
    {
      "entropy": 5.791590690612793,
      "epoch": 0.34973740517409063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991669457474559,
      "loss": 5.6514,
      "mean_token_accuracy": 0.1542787604033947,
      "num_tokens": 7810201.0,
      "step": 4495
    },
    {
      "entropy": 5.8104736328125,
      "epoch": 0.3501264345458082,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991645612768461,
      "loss": 5.5489,
      "mean_token_accuracy": 0.16673122942447663,
      "num_tokens": 7818529.0,
      "step": 4500
    },
    {
      "entropy": 5.799929141998291,
      "epoch": 0.35051546391752575,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991621734048958,
      "loss": 5.5839,
      "mean_token_accuracy": 0.16396935433149337,
      "num_tokens": 7827254.0,
      "step": 4505
    },
    {
      "entropy": 5.83071002960205,
      "epoch": 0.35090449328924334,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991597821316415,
      "loss": 5.608,
      "mean_token_accuracy": 0.1583294004201889,
      "num_tokens": 7836437.0,
      "step": 4510
    },
    {
      "entropy": 5.793945264816284,
      "epoch": 0.3512935226609609,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991573874571192,
      "loss": 5.6137,
      "mean_token_accuracy": 0.15833433270454406,
      "num_tokens": 7844939.0,
      "step": 4515
    },
    {
      "entropy": 5.79504017829895,
      "epoch": 0.35168255203267845,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991549893813653,
      "loss": 5.5967,
      "mean_token_accuracy": 0.15493495315313338,
      "num_tokens": 7852889.0,
      "step": 4520
    },
    {
      "entropy": 5.8769055843353275,
      "epoch": 0.35207158140439604,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991525879044162,
      "loss": 5.7321,
      "mean_token_accuracy": 0.14976422488689423,
      "num_tokens": 7861867.0,
      "step": 4525
    },
    {
      "entropy": 5.838841772079467,
      "epoch": 0.35246061077611357,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991501830263084,
      "loss": 5.52,
      "mean_token_accuracy": 0.16773146092891694,
      "num_tokens": 7869343.0,
      "step": 4530
    },
    {
      "entropy": 5.817440748214722,
      "epoch": 0.35284964014783116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991477747470783,
      "loss": 5.5853,
      "mean_token_accuracy": 0.1616874024271965,
      "num_tokens": 7877142.0,
      "step": 4535
    },
    {
      "entropy": 5.794790029525757,
      "epoch": 0.35323866951954874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991453630667625,
      "loss": 5.5992,
      "mean_token_accuracy": 0.16007129848003387,
      "num_tokens": 7885183.0,
      "step": 4540
    },
    {
      "entropy": 5.744523620605468,
      "epoch": 0.3536276988912663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991429479853976,
      "loss": 5.4446,
      "mean_token_accuracy": 0.16585716009140014,
      "num_tokens": 7893848.0,
      "step": 4545
    },
    {
      "entropy": 5.817834186553955,
      "epoch": 0.35401672826298386,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991405295030202,
      "loss": 5.6162,
      "mean_token_accuracy": 0.1583704322576523,
      "num_tokens": 7902438.0,
      "step": 4550
    },
    {
      "entropy": 5.83809986114502,
      "epoch": 0.35440575763470145,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991381076196671,
      "loss": 5.5867,
      "mean_token_accuracy": 0.15249855667352677,
      "num_tokens": 7910951.0,
      "step": 4555
    },
    {
      "entropy": 5.805512142181397,
      "epoch": 0.354794787006419,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991356823353748,
      "loss": 5.6928,
      "mean_token_accuracy": 0.15596385374665261,
      "num_tokens": 7920375.0,
      "step": 4560
    },
    {
      "entropy": 5.743765020370484,
      "epoch": 0.35518381637813656,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991332536501804,
      "loss": 5.5604,
      "mean_token_accuracy": 0.163666296005249,
      "num_tokens": 7928656.0,
      "step": 4565
    },
    {
      "entropy": 5.942735815048218,
      "epoch": 0.3555728457498541,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991308215641205,
      "loss": 5.7981,
      "mean_token_accuracy": 0.1519551046192646,
      "num_tokens": 7936929.0,
      "step": 4570
    },
    {
      "entropy": 5.878602027893066,
      "epoch": 0.3559618751215717,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991283860772323,
      "loss": 5.6049,
      "mean_token_accuracy": 0.1606078952550888,
      "num_tokens": 7945592.0,
      "step": 4575
    },
    {
      "entropy": 5.796780395507812,
      "epoch": 0.35635090449328927,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991259471895525,
      "loss": 5.5585,
      "mean_token_accuracy": 0.16777338832616806,
      "num_tokens": 7953656.0,
      "step": 4580
    },
    {
      "entropy": 5.812623834609985,
      "epoch": 0.3567399338650068,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991235049011182,
      "loss": 5.6175,
      "mean_token_accuracy": 0.16233777850866318,
      "num_tokens": 7962038.0,
      "step": 4585
    },
    {
      "entropy": 5.860838794708252,
      "epoch": 0.3571289632367244,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991210592119663,
      "loss": 5.7002,
      "mean_token_accuracy": 0.15201145857572557,
      "num_tokens": 7970338.0,
      "step": 4590
    },
    {
      "entropy": 5.885780715942383,
      "epoch": 0.3575179926084419,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991186101221342,
      "loss": 5.636,
      "mean_token_accuracy": 0.15820997059345246,
      "num_tokens": 7979173.0,
      "step": 4595
    },
    {
      "entropy": 5.7651495933532715,
      "epoch": 0.3579070219801595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499116157631659,
      "loss": 5.5067,
      "mean_token_accuracy": 0.16434233039617538,
      "num_tokens": 7987245.0,
      "step": 4600
    },
    {
      "entropy": 5.825799989700317,
      "epoch": 0.3582960513518771,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991137017405777,
      "loss": 5.7191,
      "mean_token_accuracy": 0.14902633428573608,
      "num_tokens": 7995551.0,
      "step": 4605
    },
    {
      "entropy": 5.7694117546081545,
      "epoch": 0.3586850807235946,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991112424489277,
      "loss": 5.5095,
      "mean_token_accuracy": 0.15749889314174653,
      "num_tokens": 8004069.0,
      "step": 4610
    },
    {
      "entropy": 5.7744873046875,
      "epoch": 0.3590741100953122,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991087797567462,
      "loss": 5.6001,
      "mean_token_accuracy": 0.16079925149679183,
      "num_tokens": 8012537.0,
      "step": 4615
    },
    {
      "entropy": 5.862816381454468,
      "epoch": 0.35946313946702974,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991063136640709,
      "loss": 5.5764,
      "mean_token_accuracy": 0.1583765998482704,
      "num_tokens": 8021387.0,
      "step": 4620
    },
    {
      "entropy": 5.805561590194702,
      "epoch": 0.3598521688387473,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991038441709388,
      "loss": 5.5482,
      "mean_token_accuracy": 0.15771045833826064,
      "num_tokens": 8030261.0,
      "step": 4625
    },
    {
      "entropy": 5.834992790222168,
      "epoch": 0.3602411982104649,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991013712773877,
      "loss": 5.6058,
      "mean_token_accuracy": 0.1562664642930031,
      "num_tokens": 8038161.0,
      "step": 4630
    },
    {
      "entropy": 5.757194852828979,
      "epoch": 0.36063022758218244,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990988949834549,
      "loss": 5.6823,
      "mean_token_accuracy": 0.16051515340805053,
      "num_tokens": 8046340.0,
      "step": 4635
    },
    {
      "entropy": 5.86285080909729,
      "epoch": 0.3610192569539,
      "grad_norm": 1.0,
      "learning_rate": 0.000499096415289178,
      "loss": 5.613,
      "mean_token_accuracy": 0.15959222316741944,
      "num_tokens": 8054997.0,
      "step": 4640
    },
    {
      "entropy": 5.812958002090454,
      "epoch": 0.3614082863256176,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990939321945948,
      "loss": 5.5542,
      "mean_token_accuracy": 0.1657189205288887,
      "num_tokens": 8063686.0,
      "step": 4645
    },
    {
      "entropy": 5.787630796432495,
      "epoch": 0.36179731569733514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990914456997427,
      "loss": 5.6709,
      "mean_token_accuracy": 0.15417664647102355,
      "num_tokens": 8072999.0,
      "step": 4650
    },
    {
      "entropy": 5.8062827587127686,
      "epoch": 0.36218634506905273,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990889558046598,
      "loss": 5.5158,
      "mean_token_accuracy": 0.16305909901857377,
      "num_tokens": 8081596.0,
      "step": 4655
    },
    {
      "entropy": 5.841017961502075,
      "epoch": 0.36257537444077026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990864625093836,
      "loss": 5.7452,
      "mean_token_accuracy": 0.15095900148153304,
      "num_tokens": 8090948.0,
      "step": 4660
    },
    {
      "entropy": 5.812306880950928,
      "epoch": 0.36296440381248785,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990839658139519,
      "loss": 5.7647,
      "mean_token_accuracy": 0.1529906287789345,
      "num_tokens": 8099445.0,
      "step": 4665
    },
    {
      "entropy": 5.870641756057739,
      "epoch": 0.36335343318420543,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990814657184028,
      "loss": 5.6048,
      "mean_token_accuracy": 0.15444873422384262,
      "num_tokens": 8108119.0,
      "step": 4670
    },
    {
      "entropy": 5.8118133544921875,
      "epoch": 0.36374246255592296,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990789622227741,
      "loss": 5.6007,
      "mean_token_accuracy": 0.15744256973266602,
      "num_tokens": 8117060.0,
      "step": 4675
    },
    {
      "entropy": 5.768152666091919,
      "epoch": 0.36413149192764055,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990764553271038,
      "loss": 5.5953,
      "mean_token_accuracy": 0.15903834998607635,
      "num_tokens": 8126229.0,
      "step": 4680
    },
    {
      "entropy": 5.848565006256104,
      "epoch": 0.3645205212993581,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990739450314299,
      "loss": 5.5379,
      "mean_token_accuracy": 0.16861870139837265,
      "num_tokens": 8134694.0,
      "step": 4685
    },
    {
      "entropy": 5.8330103874206545,
      "epoch": 0.36490955067107567,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990714313357906,
      "loss": 5.6225,
      "mean_token_accuracy": 0.16065707802772522,
      "num_tokens": 8143329.0,
      "step": 4690
    },
    {
      "entropy": 5.821296739578247,
      "epoch": 0.36529858004279325,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499068914240224,
      "loss": 5.6211,
      "mean_token_accuracy": 0.15980628728866578,
      "num_tokens": 8151432.0,
      "step": 4695
    },
    {
      "entropy": 5.806797456741333,
      "epoch": 0.3656876094145108,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990663937447682,
      "loss": 5.5666,
      "mean_token_accuracy": 0.1600451037287712,
      "num_tokens": 8159677.0,
      "step": 4700
    },
    {
      "entropy": 5.7646231174469,
      "epoch": 0.36607663878622837,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990638698494615,
      "loss": 5.6328,
      "mean_token_accuracy": 0.15402160137891768,
      "num_tokens": 8168710.0,
      "step": 4705
    },
    {
      "entropy": 5.82011308670044,
      "epoch": 0.3664656681579459,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990613425543423,
      "loss": 5.6227,
      "mean_token_accuracy": 0.16317517608404158,
      "num_tokens": 8177501.0,
      "step": 4710
    },
    {
      "entropy": 5.908048391342163,
      "epoch": 0.3668546975296635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990588118594487,
      "loss": 5.7074,
      "mean_token_accuracy": 0.14999014362692834,
      "num_tokens": 8185791.0,
      "step": 4715
    },
    {
      "entropy": 5.888072299957275,
      "epoch": 0.3672437269013811,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990562777648194,
      "loss": 5.7195,
      "mean_token_accuracy": 0.15327149331569673,
      "num_tokens": 8195020.0,
      "step": 4720
    },
    {
      "entropy": 5.806322479248047,
      "epoch": 0.3676327562730986,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990537402704928,
      "loss": 5.5863,
      "mean_token_accuracy": 0.15755437463521957,
      "num_tokens": 8203540.0,
      "step": 4725
    },
    {
      "entropy": 5.828981971740722,
      "epoch": 0.3680217856448162,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990511993765071,
      "loss": 5.5701,
      "mean_token_accuracy": 0.156550545245409,
      "num_tokens": 8211985.0,
      "step": 4730
    },
    {
      "entropy": 5.758098030090332,
      "epoch": 0.3684108150165337,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990486550829011,
      "loss": 5.6586,
      "mean_token_accuracy": 0.16203552186489106,
      "num_tokens": 8220289.0,
      "step": 4735
    },
    {
      "entropy": 5.9224365234375,
      "epoch": 0.3687998443882513,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990461073897134,
      "loss": 5.6811,
      "mean_token_accuracy": 0.1498807080090046,
      "num_tokens": 8229652.0,
      "step": 4740
    },
    {
      "entropy": 5.893592071533203,
      "epoch": 0.3691888737599689,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990435562969827,
      "loss": 5.6776,
      "mean_token_accuracy": 0.15133711099624633,
      "num_tokens": 8237728.0,
      "step": 4745
    },
    {
      "entropy": 5.796733331680298,
      "epoch": 0.3695779031316864,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990410018047475,
      "loss": 5.5632,
      "mean_token_accuracy": 0.16062172502279282,
      "num_tokens": 8246393.0,
      "step": 4750
    },
    {
      "entropy": 5.721320009231567,
      "epoch": 0.369966932503404,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990384439130467,
      "loss": 5.522,
      "mean_token_accuracy": 0.16248914897441863,
      "num_tokens": 8256736.0,
      "step": 4755
    },
    {
      "entropy": 5.824112367630005,
      "epoch": 0.3703559618751216,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990358826219192,
      "loss": 5.6279,
      "mean_token_accuracy": 0.1571101188659668,
      "num_tokens": 8265126.0,
      "step": 4760
    },
    {
      "entropy": 5.820573806762695,
      "epoch": 0.37074499124683913,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990333179314038,
      "loss": 5.5196,
      "mean_token_accuracy": 0.16383594498038292,
      "num_tokens": 8273136.0,
      "step": 4765
    },
    {
      "entropy": 5.6639457702636715,
      "epoch": 0.3711340206185567,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990307498415393,
      "loss": 5.4621,
      "mean_token_accuracy": 0.16663613840937613,
      "num_tokens": 8281697.0,
      "step": 4770
    },
    {
      "entropy": 5.7683123588562015,
      "epoch": 0.37152304999027425,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990281783523648,
      "loss": 5.5825,
      "mean_token_accuracy": 0.16584899127483368,
      "num_tokens": 8290765.0,
      "step": 4775
    },
    {
      "entropy": 5.872918605804443,
      "epoch": 0.37191207936199183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990256034639192,
      "loss": 5.6967,
      "mean_token_accuracy": 0.15300136804580688,
      "num_tokens": 8299875.0,
      "step": 4780
    },
    {
      "entropy": 5.790566539764404,
      "epoch": 0.3723011087337094,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990230251762418,
      "loss": 5.5537,
      "mean_token_accuracy": 0.1634401187300682,
      "num_tokens": 8308515.0,
      "step": 4785
    },
    {
      "entropy": 5.728052425384521,
      "epoch": 0.37269013810542695,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990204434893713,
      "loss": 5.4904,
      "mean_token_accuracy": 0.1691850394010544,
      "num_tokens": 8317260.0,
      "step": 4790
    },
    {
      "entropy": 5.847735643386841,
      "epoch": 0.37307916747714454,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990178584033474,
      "loss": 5.6432,
      "mean_token_accuracy": 0.1535599336028099,
      "num_tokens": 8325887.0,
      "step": 4795
    },
    {
      "entropy": 5.883787012100219,
      "epoch": 0.37346819684886207,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990152699182089,
      "loss": 5.6447,
      "mean_token_accuracy": 0.15114298537373544,
      "num_tokens": 8333974.0,
      "step": 4800
    },
    {
      "entropy": 5.774260997772217,
      "epoch": 0.37385722622057965,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990126780339953,
      "loss": 5.6218,
      "mean_token_accuracy": 0.15484444797039032,
      "num_tokens": 8343199.0,
      "step": 4805
    },
    {
      "entropy": 5.702182579040527,
      "epoch": 0.37424625559229724,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990100827507459,
      "loss": 5.4226,
      "mean_token_accuracy": 0.16433848440647125,
      "num_tokens": 8351202.0,
      "step": 4810
    },
    {
      "entropy": 5.862250757217407,
      "epoch": 0.37463528496401477,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990074840684999,
      "loss": 5.7634,
      "mean_token_accuracy": 0.1481994055211544,
      "num_tokens": 8360011.0,
      "step": 4815
    },
    {
      "entropy": 5.797674608230591,
      "epoch": 0.37502431433573236,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990048819872969,
      "loss": 5.6219,
      "mean_token_accuracy": 0.16221669912338257,
      "num_tokens": 8367874.0,
      "step": 4820
    },
    {
      "entropy": 5.823427486419678,
      "epoch": 0.3754133437074499,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990022765071765,
      "loss": 5.5947,
      "mean_token_accuracy": 0.1601307660341263,
      "num_tokens": 8376708.0,
      "step": 4825
    },
    {
      "entropy": 5.858452033996582,
      "epoch": 0.3758023730791675,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498999667628178,
      "loss": 5.6088,
      "mean_token_accuracy": 0.15857539996504783,
      "num_tokens": 8385467.0,
      "step": 4830
    },
    {
      "entropy": 5.77020845413208,
      "epoch": 0.37619140245088506,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989970553503411,
      "loss": 5.5836,
      "mean_token_accuracy": 0.15638737380504608,
      "num_tokens": 8393574.0,
      "step": 4835
    },
    {
      "entropy": 5.769679689407349,
      "epoch": 0.3765804318226026,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989944396737054,
      "loss": 5.625,
      "mean_token_accuracy": 0.15962510854005812,
      "num_tokens": 8402857.0,
      "step": 4840
    },
    {
      "entropy": 5.779857015609741,
      "epoch": 0.3769694611943202,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989918205983106,
      "loss": 5.635,
      "mean_token_accuracy": 0.1616414651274681,
      "num_tokens": 8411728.0,
      "step": 4845
    },
    {
      "entropy": 5.835721397399903,
      "epoch": 0.37735849056603776,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989891981241964,
      "loss": 5.6033,
      "mean_token_accuracy": 0.16248496770858764,
      "num_tokens": 8420307.0,
      "step": 4850
    },
    {
      "entropy": 5.85571141242981,
      "epoch": 0.3777475199377553,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989865722514027,
      "loss": 5.6193,
      "mean_token_accuracy": 0.15614497512578965,
      "num_tokens": 8429442.0,
      "step": 4855
    },
    {
      "entropy": 5.753107929229737,
      "epoch": 0.3781365493094729,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989839429799692,
      "loss": 5.5144,
      "mean_token_accuracy": 0.1640685573220253,
      "num_tokens": 8437945.0,
      "step": 4860
    },
    {
      "entropy": 5.734889984130859,
      "epoch": 0.3785255786811904,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498981310309936,
      "loss": 5.5018,
      "mean_token_accuracy": 0.16307622343301773,
      "num_tokens": 8445743.0,
      "step": 4865
    },
    {
      "entropy": 5.741227006912231,
      "epoch": 0.378914608052908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989786742413428,
      "loss": 5.5981,
      "mean_token_accuracy": 0.15394424200057982,
      "num_tokens": 8454091.0,
      "step": 4870
    },
    {
      "entropy": 5.900302505493164,
      "epoch": 0.3793036374246256,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989760347742298,
      "loss": 5.6053,
      "mean_token_accuracy": 0.15447922945022582,
      "num_tokens": 8461992.0,
      "step": 4875
    },
    {
      "entropy": 5.767937135696411,
      "epoch": 0.3796926667963431,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989733919086369,
      "loss": 5.5546,
      "mean_token_accuracy": 0.15787096470594406,
      "num_tokens": 8470730.0,
      "step": 4880
    },
    {
      "entropy": 5.739106273651123,
      "epoch": 0.3800816961680607,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989707456446043,
      "loss": 5.5916,
      "mean_token_accuracy": 0.15556688904762267,
      "num_tokens": 8479947.0,
      "step": 4885
    },
    {
      "entropy": 5.873720502853393,
      "epoch": 0.38047072553977823,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989680959821721,
      "loss": 5.6699,
      "mean_token_accuracy": 0.1565061554312706,
      "num_tokens": 8488237.0,
      "step": 4890
    },
    {
      "entropy": 5.827173280715942,
      "epoch": 0.3808597549114958,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989654429213805,
      "loss": 5.5369,
      "mean_token_accuracy": 0.1635414183139801,
      "num_tokens": 8497146.0,
      "step": 4895
    },
    {
      "entropy": 5.802580404281616,
      "epoch": 0.3812487842832134,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989627864622699,
      "loss": 5.6021,
      "mean_token_accuracy": 0.1608035609126091,
      "num_tokens": 8505736.0,
      "step": 4900
    },
    {
      "entropy": 5.807352447509766,
      "epoch": 0.38163781365493094,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989601266048804,
      "loss": 5.622,
      "mean_token_accuracy": 0.15869393199682236,
      "num_tokens": 8515368.0,
      "step": 4905
    },
    {
      "entropy": 5.755753421783448,
      "epoch": 0.3820268430266485,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989574633492525,
      "loss": 5.6008,
      "mean_token_accuracy": 0.15789534896612167,
      "num_tokens": 8524141.0,
      "step": 4910
    },
    {
      "entropy": 5.77313814163208,
      "epoch": 0.38241587239836605,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989547966954266,
      "loss": 5.5015,
      "mean_token_accuracy": 0.16294147670269013,
      "num_tokens": 8533116.0,
      "step": 4915
    },
    {
      "entropy": 5.801529884338379,
      "epoch": 0.38280490177008364,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498952126643443,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1634344071149826,
      "num_tokens": 8541919.0,
      "step": 4920
    },
    {
      "entropy": 5.777447128295899,
      "epoch": 0.3831939311418012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989494531933424,
      "loss": 5.528,
      "mean_token_accuracy": 0.16603374779224395,
      "num_tokens": 8550372.0,
      "step": 4925
    },
    {
      "entropy": 5.7857917785644535,
      "epoch": 0.38358296051351876,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989467763451652,
      "loss": 5.6173,
      "mean_token_accuracy": 0.15723662078380585,
      "num_tokens": 8559246.0,
      "step": 4930
    },
    {
      "entropy": 5.734502744674683,
      "epoch": 0.38397198988523634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989440960989523,
      "loss": 5.3871,
      "mean_token_accuracy": 0.16804315894842148,
      "num_tokens": 8567005.0,
      "step": 4935
    },
    {
      "entropy": 5.719801902770996,
      "epoch": 0.3843610192569539,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498941412454744,
      "loss": 5.5876,
      "mean_token_accuracy": 0.15178390145301818,
      "num_tokens": 8575399.0,
      "step": 4940
    },
    {
      "entropy": 5.837953519821167,
      "epoch": 0.38475004862867146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989387254125813,
      "loss": 5.6603,
      "mean_token_accuracy": 0.15484791696071626,
      "num_tokens": 8583994.0,
      "step": 4945
    },
    {
      "entropy": 5.754817771911621,
      "epoch": 0.38513907800038905,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989360349725049,
      "loss": 5.5737,
      "mean_token_accuracy": 0.16358229368925095,
      "num_tokens": 8593446.0,
      "step": 4950
    },
    {
      "entropy": 5.788543653488159,
      "epoch": 0.3855281073721066,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989333411345555,
      "loss": 5.5801,
      "mean_token_accuracy": 0.165724715590477,
      "num_tokens": 8602821.0,
      "step": 4955
    },
    {
      "entropy": 5.805869770050049,
      "epoch": 0.38591713674382416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989306438987742,
      "loss": 5.6849,
      "mean_token_accuracy": 0.15382945835590361,
      "num_tokens": 8611831.0,
      "step": 4960
    },
    {
      "entropy": 5.870657587051392,
      "epoch": 0.38630616611554175,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989279432652018,
      "loss": 5.5864,
      "mean_token_accuracy": 0.15871587842702867,
      "num_tokens": 8620986.0,
      "step": 4965
    },
    {
      "entropy": 5.799157762527466,
      "epoch": 0.3866951954872593,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989252392338791,
      "loss": 5.515,
      "mean_token_accuracy": 0.16157624498009682,
      "num_tokens": 8629822.0,
      "step": 4970
    },
    {
      "entropy": 5.760199642181396,
      "epoch": 0.38708422485897687,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989225318048475,
      "loss": 5.6182,
      "mean_token_accuracy": 0.15575948357582092,
      "num_tokens": 8637679.0,
      "step": 4975
    },
    {
      "entropy": 5.85200686454773,
      "epoch": 0.3874732542306944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989198209781478,
      "loss": 5.647,
      "mean_token_accuracy": 0.1581929624080658,
      "num_tokens": 8646705.0,
      "step": 4980
    },
    {
      "entropy": 5.861097717285157,
      "epoch": 0.387862283602412,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989171067538211,
      "loss": 5.5706,
      "mean_token_accuracy": 0.15901409238576888,
      "num_tokens": 8655599.0,
      "step": 4985
    },
    {
      "entropy": 5.733076524734497,
      "epoch": 0.38825131297412957,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498914389131909,
      "loss": 5.5962,
      "mean_token_accuracy": 0.15992813110351561,
      "num_tokens": 8665025.0,
      "step": 4990
    },
    {
      "entropy": 5.786661195755005,
      "epoch": 0.3886403423458471,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989116681124523,
      "loss": 5.4636,
      "mean_token_accuracy": 0.1661472424864769,
      "num_tokens": 8672994.0,
      "step": 4995
    },
    {
      "entropy": 5.734365653991699,
      "epoch": 0.3890293717175647,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989089436954924,
      "loss": 5.6162,
      "mean_token_accuracy": 0.15561984330415726,
      "num_tokens": 8682457.0,
      "step": 5000
    },
    {
      "entropy": 5.826134824752808,
      "epoch": 0.3894184010892822,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989062158810706,
      "loss": 5.5833,
      "mean_token_accuracy": 0.16018826365470887,
      "num_tokens": 8690888.0,
      "step": 5005
    },
    {
      "entropy": 5.903531408309936,
      "epoch": 0.3898074304609998,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989034846692284,
      "loss": 5.6732,
      "mean_token_accuracy": 0.15570265799760818,
      "num_tokens": 8699632.0,
      "step": 5010
    },
    {
      "entropy": 5.701883125305176,
      "epoch": 0.3901964598327174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989007500600073,
      "loss": 5.5154,
      "mean_token_accuracy": 0.16052396595478058,
      "num_tokens": 8707955.0,
      "step": 5015
    },
    {
      "entropy": 5.798250961303711,
      "epoch": 0.3905854892044349,
      "grad_norm": 1.125,
      "learning_rate": 0.0004988980120534486,
      "loss": 5.6663,
      "mean_token_accuracy": 0.16303107291460037,
      "num_tokens": 8716103.0,
      "step": 5020
    },
    {
      "entropy": 5.777353525161743,
      "epoch": 0.3909745185761525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498895270649594,
      "loss": 5.5408,
      "mean_token_accuracy": 0.168153877556324,
      "num_tokens": 8724499.0,
      "step": 5025
    },
    {
      "entropy": 5.714413070678711,
      "epoch": 0.39136354794787004,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498892525848485,
      "loss": 5.534,
      "mean_token_accuracy": 0.16576792299747467,
      "num_tokens": 8732829.0,
      "step": 5030
    },
    {
      "entropy": 5.708601713180542,
      "epoch": 0.3917525773195876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988897776501633,
      "loss": 5.4153,
      "mean_token_accuracy": 0.1727154091000557,
      "num_tokens": 8741617.0,
      "step": 5035
    },
    {
      "entropy": 5.699780130386353,
      "epoch": 0.3921416066913052,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988870260546705,
      "loss": 5.4005,
      "mean_token_accuracy": 0.16900650709867476,
      "num_tokens": 8750146.0,
      "step": 5040
    },
    {
      "entropy": 5.752245140075684,
      "epoch": 0.39253063606302274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988842710620486,
      "loss": 5.5554,
      "mean_token_accuracy": 0.1608773484826088,
      "num_tokens": 8758744.0,
      "step": 5045
    },
    {
      "entropy": 5.730192708969116,
      "epoch": 0.39291966543474033,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988815126723391,
      "loss": 5.594,
      "mean_token_accuracy": 0.16051646322011948,
      "num_tokens": 8768219.0,
      "step": 5050
    },
    {
      "entropy": 5.803083086013794,
      "epoch": 0.3933086948064579,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988787508855841,
      "loss": 5.5881,
      "mean_token_accuracy": 0.15734401047229768,
      "num_tokens": 8777047.0,
      "step": 5055
    },
    {
      "entropy": 5.740478944778443,
      "epoch": 0.39369772417817545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988759857018253,
      "loss": 5.4748,
      "mean_token_accuracy": 0.16194416731595992,
      "num_tokens": 8785918.0,
      "step": 5060
    },
    {
      "entropy": 5.72450623512268,
      "epoch": 0.39408675354989303,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988732171211048,
      "loss": 5.5592,
      "mean_token_accuracy": 0.15556975752115249,
      "num_tokens": 8794253.0,
      "step": 5065
    },
    {
      "entropy": 5.776522016525268,
      "epoch": 0.39447578292161056,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988704451434644,
      "loss": 5.6091,
      "mean_token_accuracy": 0.16497571170330047,
      "num_tokens": 8802352.0,
      "step": 5070
    },
    {
      "entropy": 5.831839084625244,
      "epoch": 0.39486481229332815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988676697689465,
      "loss": 5.5726,
      "mean_token_accuracy": 0.1600097894668579,
      "num_tokens": 8810956.0,
      "step": 5075
    },
    {
      "entropy": 5.660107612609863,
      "epoch": 0.39525384166504574,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498864890997593,
      "loss": 5.483,
      "mean_token_accuracy": 0.16221728324890136,
      "num_tokens": 8819062.0,
      "step": 5080
    },
    {
      "entropy": 5.754838228225708,
      "epoch": 0.39564287103676327,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988621088294461,
      "loss": 5.5921,
      "mean_token_accuracy": 0.15481371581554412,
      "num_tokens": 8827470.0,
      "step": 5085
    },
    {
      "entropy": 5.7890125751495365,
      "epoch": 0.39603190040848085,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498859323264548,
      "loss": 5.6118,
      "mean_token_accuracy": 0.15711697936058044,
      "num_tokens": 8837003.0,
      "step": 5090
    },
    {
      "entropy": 5.8137959957122805,
      "epoch": 0.3964209297801984,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498856534302941,
      "loss": 5.622,
      "mean_token_accuracy": 0.15432859659194947,
      "num_tokens": 8846605.0,
      "step": 5095
    },
    {
      "entropy": 5.822982692718506,
      "epoch": 0.39680995915191597,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988537419446673,
      "loss": 5.6472,
      "mean_token_accuracy": 0.15490608364343644,
      "num_tokens": 8855232.0,
      "step": 5100
    },
    {
      "entropy": 5.805551242828369,
      "epoch": 0.39719898852363356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988509461897694,
      "loss": 5.5391,
      "mean_token_accuracy": 0.163739974796772,
      "num_tokens": 8864176.0,
      "step": 5105
    },
    {
      "entropy": 5.79059329032898,
      "epoch": 0.3975880178953511,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988481470382897,
      "loss": 5.6013,
      "mean_token_accuracy": 0.16403751224279403,
      "num_tokens": 8872671.0,
      "step": 5110
    },
    {
      "entropy": 5.662219524383545,
      "epoch": 0.3979770472670687,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988453444902708,
      "loss": 5.5092,
      "mean_token_accuracy": 0.1657335177063942,
      "num_tokens": 8881316.0,
      "step": 5115
    },
    {
      "entropy": 5.828357601165772,
      "epoch": 0.3983660766387862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988425385457549,
      "loss": 5.6452,
      "mean_token_accuracy": 0.16099053174257277,
      "num_tokens": 8890114.0,
      "step": 5120
    },
    {
      "entropy": 5.7877702713012695,
      "epoch": 0.3987551060105038,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004988397292047848,
      "loss": 5.564,
      "mean_token_accuracy": 0.16202669739723205,
      "num_tokens": 8899598.0,
      "step": 5125
    },
    {
      "entropy": 5.643172311782837,
      "epoch": 0.3991441353822214,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988369164674032,
      "loss": 5.4223,
      "mean_token_accuracy": 0.16724593490362166,
      "num_tokens": 8908244.0,
      "step": 5130
    },
    {
      "entropy": 5.731029653549195,
      "epoch": 0.3995331647539389,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988341003336526,
      "loss": 5.4788,
      "mean_token_accuracy": 0.16071776747703553,
      "num_tokens": 8916280.0,
      "step": 5135
    },
    {
      "entropy": 5.814835882186889,
      "epoch": 0.3999221941256565,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988312808035757,
      "loss": 5.5257,
      "mean_token_accuracy": 0.15995677262544633,
      "num_tokens": 8923947.0,
      "step": 5140
    },
    {
      "entropy": 5.729617071151734,
      "epoch": 0.400311223497374,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988284578772155,
      "loss": 5.5708,
      "mean_token_accuracy": 0.15164806991815566,
      "num_tokens": 8933374.0,
      "step": 5145
    },
    {
      "entropy": 5.724401140213013,
      "epoch": 0.4007002528690916,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988256315546146,
      "loss": 5.5003,
      "mean_token_accuracy": 0.1619577333331108,
      "num_tokens": 8941757.0,
      "step": 5150
    },
    {
      "entropy": 5.732602739334107,
      "epoch": 0.4010892822408092,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988228018358161,
      "loss": 5.4125,
      "mean_token_accuracy": 0.16584188491106033,
      "num_tokens": 8950128.0,
      "step": 5155
    },
    {
      "entropy": 5.754231929779053,
      "epoch": 0.40147831161252673,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988199687208628,
      "loss": 5.5701,
      "mean_token_accuracy": 0.15558881908655167,
      "num_tokens": 8958939.0,
      "step": 5160
    },
    {
      "entropy": 5.791546297073364,
      "epoch": 0.4018673409842443,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988171322097977,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16313695162534714,
      "num_tokens": 8967519.0,
      "step": 5165
    },
    {
      "entropy": 5.655329084396362,
      "epoch": 0.4022563703559619,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988142923026638,
      "loss": 5.3946,
      "mean_token_accuracy": 0.1725483238697052,
      "num_tokens": 8976215.0,
      "step": 5170
    },
    {
      "entropy": 5.770650625228882,
      "epoch": 0.40264539972767943,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988114489995044,
      "loss": 5.6178,
      "mean_token_accuracy": 0.1560276210308075,
      "num_tokens": 8984060.0,
      "step": 5175
    },
    {
      "entropy": 5.828380680084228,
      "epoch": 0.403034429099397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988086023003624,
      "loss": 5.5662,
      "mean_token_accuracy": 0.1560399517416954,
      "num_tokens": 8992301.0,
      "step": 5180
    },
    {
      "entropy": 5.813635921478271,
      "epoch": 0.40342345847111455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988057522052811,
      "loss": 5.5755,
      "mean_token_accuracy": 0.161184823513031,
      "num_tokens": 9000517.0,
      "step": 5185
    },
    {
      "entropy": 5.704269933700561,
      "epoch": 0.40381248784283214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988028987143037,
      "loss": 5.5563,
      "mean_token_accuracy": 0.15821606069803237,
      "num_tokens": 9008826.0,
      "step": 5190
    },
    {
      "entropy": 5.7354576110839846,
      "epoch": 0.4042015172145497,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988000418274736,
      "loss": 5.5362,
      "mean_token_accuracy": 0.15350192189216613,
      "num_tokens": 9017952.0,
      "step": 5195
    },
    {
      "entropy": 5.7346971988677975,
      "epoch": 0.40459054658626725,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987971815448341,
      "loss": 5.5303,
      "mean_token_accuracy": 0.16944959610700608,
      "num_tokens": 9026688.0,
      "step": 5200
    },
    {
      "entropy": 5.783084297180176,
      "epoch": 0.40497957595798484,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987943178664285,
      "loss": 5.5106,
      "mean_token_accuracy": 0.1653973266482353,
      "num_tokens": 9035761.0,
      "step": 5205
    },
    {
      "entropy": 5.7801374912261965,
      "epoch": 0.40536860532970237,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987914507923004,
      "loss": 5.6219,
      "mean_token_accuracy": 0.15662121027708054,
      "num_tokens": 9044510.0,
      "step": 5210
    },
    {
      "entropy": 5.7042594909667965,
      "epoch": 0.40575763470141996,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987885803224931,
      "loss": 5.488,
      "mean_token_accuracy": 0.16024671047925948,
      "num_tokens": 9053031.0,
      "step": 5215
    },
    {
      "entropy": 5.6758147239685055,
      "epoch": 0.40614666407313754,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987857064570505,
      "loss": 5.4703,
      "mean_token_accuracy": 0.16776451021432875,
      "num_tokens": 9061580.0,
      "step": 5220
    },
    {
      "entropy": 5.8110490322113035,
      "epoch": 0.4065356934448551,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987828291960158,
      "loss": 5.5996,
      "mean_token_accuracy": 0.1625698685646057,
      "num_tokens": 9069741.0,
      "step": 5225
    },
    {
      "entropy": 5.859944915771484,
      "epoch": 0.40692472281657266,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498779948539433,
      "loss": 5.6224,
      "mean_token_accuracy": 0.1524021878838539,
      "num_tokens": 9079214.0,
      "step": 5230
    },
    {
      "entropy": 5.8057465076446535,
      "epoch": 0.4073137521882902,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987770644873455,
      "loss": 5.5685,
      "mean_token_accuracy": 0.1614005982875824,
      "num_tokens": 9088201.0,
      "step": 5235
    },
    {
      "entropy": 5.779406118392944,
      "epoch": 0.4077027815600078,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987741770397974,
      "loss": 5.6117,
      "mean_token_accuracy": 0.15523144751787185,
      "num_tokens": 9096829.0,
      "step": 5240
    },
    {
      "entropy": 5.743911600112915,
      "epoch": 0.40809181093172536,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987712861968323,
      "loss": 5.5435,
      "mean_token_accuracy": 0.1675190418958664,
      "num_tokens": 9105515.0,
      "step": 5245
    },
    {
      "entropy": 5.786427354812622,
      "epoch": 0.4084808403034429,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498768391958494,
      "loss": 5.5222,
      "mean_token_accuracy": 0.16446349769830704,
      "num_tokens": 9113614.0,
      "step": 5250
    },
    {
      "entropy": 5.735589265823364,
      "epoch": 0.4088698696751605,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987654943248266,
      "loss": 5.6039,
      "mean_token_accuracy": 0.15211761370301247,
      "num_tokens": 9122337.0,
      "step": 5255
    },
    {
      "entropy": 5.8131852626800535,
      "epoch": 0.40925889904687807,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987625932958739,
      "loss": 5.5079,
      "mean_token_accuracy": 0.1656147763133049,
      "num_tokens": 9130417.0,
      "step": 5260
    },
    {
      "entropy": 5.75497784614563,
      "epoch": 0.4096479284185956,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987596888716801,
      "loss": 5.5501,
      "mean_token_accuracy": 0.15596065521240235,
      "num_tokens": 9139317.0,
      "step": 5265
    },
    {
      "entropy": 5.73632550239563,
      "epoch": 0.4100369577903132,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987567810522892,
      "loss": 5.536,
      "mean_token_accuracy": 0.1615392506122589,
      "num_tokens": 9147808.0,
      "step": 5270
    },
    {
      "entropy": 5.884146213531494,
      "epoch": 0.4104259871620307,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987538698377451,
      "loss": 5.6492,
      "mean_token_accuracy": 0.1503283016383648,
      "num_tokens": 9157117.0,
      "step": 5275
    },
    {
      "entropy": 5.776542234420776,
      "epoch": 0.4108150165337483,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987509552280924,
      "loss": 5.4896,
      "mean_token_accuracy": 0.1617259919643402,
      "num_tokens": 9165297.0,
      "step": 5280
    },
    {
      "entropy": 5.791871118545532,
      "epoch": 0.4112040459054659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498748037223375,
      "loss": 5.6357,
      "mean_token_accuracy": 0.16000361293554305,
      "num_tokens": 9172945.0,
      "step": 5285
    },
    {
      "entropy": 5.813172388076782,
      "epoch": 0.4115930752771834,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987451158236372,
      "loss": 5.5672,
      "mean_token_accuracy": 0.16503645181655885,
      "num_tokens": 9181445.0,
      "step": 5290
    },
    {
      "entropy": 5.679479742050171,
      "epoch": 0.411982104648901,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987421910289234,
      "loss": 5.4169,
      "mean_token_accuracy": 0.16110423803329468,
      "num_tokens": 9189535.0,
      "step": 5295
    },
    {
      "entropy": 5.720636653900146,
      "epoch": 0.41237113402061853,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987392628392781,
      "loss": 5.6179,
      "mean_token_accuracy": 0.15802470296621324,
      "num_tokens": 9198328.0,
      "step": 5300
    },
    {
      "entropy": 5.82599983215332,
      "epoch": 0.4127601633923361,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987363312547456,
      "loss": 5.5101,
      "mean_token_accuracy": 0.16403015553951264,
      "num_tokens": 9206840.0,
      "step": 5305
    },
    {
      "entropy": 5.707152414321899,
      "epoch": 0.4131491927640537,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987333962753703,
      "loss": 5.4306,
      "mean_token_accuracy": 0.15987958312034606,
      "num_tokens": 9215155.0,
      "step": 5310
    },
    {
      "entropy": 5.705473613739014,
      "epoch": 0.41353822213577124,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987304579011967,
      "loss": 5.6003,
      "mean_token_accuracy": 0.16152040511369706,
      "num_tokens": 9224426.0,
      "step": 5315
    },
    {
      "entropy": 5.80241584777832,
      "epoch": 0.4139272515074888,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987275161322697,
      "loss": 5.6134,
      "mean_token_accuracy": 0.16057688891887664,
      "num_tokens": 9233785.0,
      "step": 5320
    },
    {
      "entropy": 5.7694268226623535,
      "epoch": 0.41431628087920636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987245709686337,
      "loss": 5.5674,
      "mean_token_accuracy": 0.1608361631631851,
      "num_tokens": 9242168.0,
      "step": 5325
    },
    {
      "entropy": 5.765720796585083,
      "epoch": 0.41470531025092394,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987216224103334,
      "loss": 5.5899,
      "mean_token_accuracy": 0.16129221320152282,
      "num_tokens": 9250862.0,
      "step": 5330
    },
    {
      "entropy": 5.77749605178833,
      "epoch": 0.41509433962264153,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987186704574136,
      "loss": 5.5997,
      "mean_token_accuracy": 0.1521721214056015,
      "num_tokens": 9259957.0,
      "step": 5335
    },
    {
      "entropy": 5.711634588241577,
      "epoch": 0.41548336899435906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987157151099189,
      "loss": 5.4631,
      "mean_token_accuracy": 0.16066141575574874,
      "num_tokens": 9268590.0,
      "step": 5340
    },
    {
      "entropy": 5.777009773254394,
      "epoch": 0.41587239836607665,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987127563678945,
      "loss": 5.5336,
      "mean_token_accuracy": 0.1600711613893509,
      "num_tokens": 9276872.0,
      "step": 5345
    },
    {
      "entropy": 5.714554786682129,
      "epoch": 0.4162614277377942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987097942313852,
      "loss": 5.4275,
      "mean_token_accuracy": 0.16699732095003128,
      "num_tokens": 9284556.0,
      "step": 5350
    },
    {
      "entropy": 5.701636886596679,
      "epoch": 0.41665045710951176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987068287004355,
      "loss": 5.5746,
      "mean_token_accuracy": 0.16003220081329345,
      "num_tokens": 9293315.0,
      "step": 5355
    },
    {
      "entropy": 5.799536657333374,
      "epoch": 0.41703948648122935,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987038597750909,
      "loss": 5.5539,
      "mean_token_accuracy": 0.16359138786792754,
      "num_tokens": 9302923.0,
      "step": 5360
    },
    {
      "entropy": 5.7929441928863525,
      "epoch": 0.4174285158529469,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987008874553964,
      "loss": 5.4853,
      "mean_token_accuracy": 0.1643255814909935,
      "num_tokens": 9311388.0,
      "step": 5365
    },
    {
      "entropy": 5.703044176101685,
      "epoch": 0.41781754522466447,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986979117413969,
      "loss": 5.4553,
      "mean_token_accuracy": 0.16968912333250047,
      "num_tokens": 9319880.0,
      "step": 5370
    },
    {
      "entropy": 5.727995777130127,
      "epoch": 0.41820657459638205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986949326331376,
      "loss": 5.5841,
      "mean_token_accuracy": 0.16095167100429536,
      "num_tokens": 9328374.0,
      "step": 5375
    },
    {
      "entropy": 5.86257700920105,
      "epoch": 0.4185956039680996,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986919501306638,
      "loss": 5.5299,
      "mean_token_accuracy": 0.17047778517007828,
      "num_tokens": 9336428.0,
      "step": 5380
    },
    {
      "entropy": 5.690197086334228,
      "epoch": 0.41898463333981717,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986889642340207,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1621948793530464,
      "num_tokens": 9345794.0,
      "step": 5385
    },
    {
      "entropy": 5.817267942428589,
      "epoch": 0.4193736627115347,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986859749432536,
      "loss": 5.5843,
      "mean_token_accuracy": 0.1633555069565773,
      "num_tokens": 9354724.0,
      "step": 5390
    },
    {
      "entropy": 5.847879076004029,
      "epoch": 0.4197626920832523,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986829822584078,
      "loss": 5.6044,
      "mean_token_accuracy": 0.15595599561929702,
      "num_tokens": 9364008.0,
      "step": 5395
    },
    {
      "entropy": 5.760229253768921,
      "epoch": 0.4201517214549699,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986799861795289,
      "loss": 5.5628,
      "mean_token_accuracy": 0.16657146364450454,
      "num_tokens": 9372409.0,
      "step": 5400
    },
    {
      "entropy": 5.756686019897461,
      "epoch": 0.4205407508266874,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986769867066622,
      "loss": 5.5111,
      "mean_token_accuracy": 0.1584501311182976,
      "num_tokens": 9381172.0,
      "step": 5405
    },
    {
      "entropy": 5.765036106109619,
      "epoch": 0.420929780198405,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986739838398532,
      "loss": 5.5309,
      "mean_token_accuracy": 0.16727976500988007,
      "num_tokens": 9389245.0,
      "step": 5410
    },
    {
      "entropy": 5.771451425552368,
      "epoch": 0.4213188095701225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986709775791475,
      "loss": 5.5546,
      "mean_token_accuracy": 0.16408045440912247,
      "num_tokens": 9397954.0,
      "step": 5415
    },
    {
      "entropy": 5.781752395629883,
      "epoch": 0.4217078389418401,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986679679245907,
      "loss": 5.6491,
      "mean_token_accuracy": 0.15681129544973374,
      "num_tokens": 9406986.0,
      "step": 5420
    },
    {
      "entropy": 5.76668906211853,
      "epoch": 0.4220968683135577,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986649548762286,
      "loss": 5.6079,
      "mean_token_accuracy": 0.16251776367425919,
      "num_tokens": 9415163.0,
      "step": 5425
    },
    {
      "entropy": 5.703148698806762,
      "epoch": 0.4224858976852752,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986619384341066,
      "loss": 5.4274,
      "mean_token_accuracy": 0.17236438244581223,
      "num_tokens": 9423422.0,
      "step": 5430
    },
    {
      "entropy": 5.703578805923462,
      "epoch": 0.4228749270569928,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986589185982708,
      "loss": 5.4681,
      "mean_token_accuracy": 0.17654578536748886,
      "num_tokens": 9431349.0,
      "step": 5435
    },
    {
      "entropy": 5.750101041793823,
      "epoch": 0.42326395642871034,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986558953687671,
      "loss": 5.5691,
      "mean_token_accuracy": 0.1597389116883278,
      "num_tokens": 9439715.0,
      "step": 5440
    },
    {
      "entropy": 5.782275152206421,
      "epoch": 0.42365298580042793,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986528687456409,
      "loss": 5.5734,
      "mean_token_accuracy": 0.16293269544839858,
      "num_tokens": 9448216.0,
      "step": 5445
    },
    {
      "entropy": 5.811833715438842,
      "epoch": 0.4240420151721455,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986498387289384,
      "loss": 5.5344,
      "mean_token_accuracy": 0.16137373745441436,
      "num_tokens": 9456346.0,
      "step": 5450
    },
    {
      "entropy": 5.736314725875855,
      "epoch": 0.42443104454386305,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986468053187058,
      "loss": 5.5665,
      "mean_token_accuracy": 0.16170335710048675,
      "num_tokens": 9465476.0,
      "step": 5455
    },
    {
      "entropy": 5.7815176486969,
      "epoch": 0.42482007391558063,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986437685149887,
      "loss": 5.6367,
      "mean_token_accuracy": 0.1560618445277214,
      "num_tokens": 9474582.0,
      "step": 5460
    },
    {
      "entropy": 5.797699975967407,
      "epoch": 0.4252091032872982,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986407283178334,
      "loss": 5.4995,
      "mean_token_accuracy": 0.15812324583530427,
      "num_tokens": 9482801.0,
      "step": 5465
    },
    {
      "entropy": 5.703231477737427,
      "epoch": 0.42559813265901575,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986376847272861,
      "loss": 5.5451,
      "mean_token_accuracy": 0.16197668462991716,
      "num_tokens": 9490894.0,
      "step": 5470
    },
    {
      "entropy": 5.715131330490112,
      "epoch": 0.42598716203073334,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986346377433929,
      "loss": 5.4799,
      "mean_token_accuracy": 0.1601783439517021,
      "num_tokens": 9499414.0,
      "step": 5475
    },
    {
      "entropy": 5.728883600234985,
      "epoch": 0.42637619140245087,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986315873661999,
      "loss": 5.4638,
      "mean_token_accuracy": 0.16241822242736817,
      "num_tokens": 9508945.0,
      "step": 5480
    },
    {
      "entropy": 5.7178935527801515,
      "epoch": 0.42676522077416845,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986285335957536,
      "loss": 5.5351,
      "mean_token_accuracy": 0.1633280724287033,
      "num_tokens": 9517833.0,
      "step": 5485
    },
    {
      "entropy": 5.7881711483001705,
      "epoch": 0.42715425014588604,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986254764321002,
      "loss": 5.6887,
      "mean_token_accuracy": 0.15465230494737625,
      "num_tokens": 9526551.0,
      "step": 5490
    },
    {
      "entropy": 5.733046913146973,
      "epoch": 0.42754327951760357,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986224158752861,
      "loss": 5.4539,
      "mean_token_accuracy": 0.16491978913545607,
      "num_tokens": 9535224.0,
      "step": 5495
    },
    {
      "entropy": 5.765879344940186,
      "epoch": 0.42793230888932116,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986193519253578,
      "loss": 5.5475,
      "mean_token_accuracy": 0.15739070773124694,
      "num_tokens": 9544466.0,
      "step": 5500
    },
    {
      "entropy": 5.721788835525513,
      "epoch": 0.4283213382610387,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986162845823618,
      "loss": 5.5546,
      "mean_token_accuracy": 0.16208747327327727,
      "num_tokens": 9553520.0,
      "step": 5505
    },
    {
      "entropy": 5.7257012844085695,
      "epoch": 0.4287103676327563,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986132138463446,
      "loss": 5.4707,
      "mean_token_accuracy": 0.1690576672554016,
      "num_tokens": 9562053.0,
      "step": 5510
    },
    {
      "entropy": 5.719926404953003,
      "epoch": 0.42909939700447386,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986101397173528,
      "loss": 5.4614,
      "mean_token_accuracy": 0.16765471994876863,
      "num_tokens": 9570632.0,
      "step": 5515
    },
    {
      "entropy": 5.755585145950318,
      "epoch": 0.4294884263761914,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498607062195433,
      "loss": 5.4611,
      "mean_token_accuracy": 0.1679205983877182,
      "num_tokens": 9578739.0,
      "step": 5520
    },
    {
      "entropy": 5.68006854057312,
      "epoch": 0.429877455747909,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498603981280632,
      "loss": 5.5014,
      "mean_token_accuracy": 0.16100622415542604,
      "num_tokens": 9587314.0,
      "step": 5525
    },
    {
      "entropy": 5.749175643920898,
      "epoch": 0.4302664851196265,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986008969729964,
      "loss": 5.5621,
      "mean_token_accuracy": 0.1643988236784935,
      "num_tokens": 9595954.0,
      "step": 5530
    },
    {
      "entropy": 5.813189172744751,
      "epoch": 0.4306555144913441,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985978092725731,
      "loss": 5.5733,
      "mean_token_accuracy": 0.15979394167661667,
      "num_tokens": 9604767.0,
      "step": 5535
    },
    {
      "entropy": 5.6896750926971436,
      "epoch": 0.4310445438630617,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498594718179409,
      "loss": 5.4301,
      "mean_token_accuracy": 0.16427869349718094,
      "num_tokens": 9613173.0,
      "step": 5540
    },
    {
      "entropy": 5.789145374298096,
      "epoch": 0.4314335732347792,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985916236935508,
      "loss": 5.5244,
      "mean_token_accuracy": 0.16085876896977425,
      "num_tokens": 9622777.0,
      "step": 5545
    },
    {
      "entropy": 5.71923565864563,
      "epoch": 0.4318226026064968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985885258150458,
      "loss": 5.4801,
      "mean_token_accuracy": 0.15869101583957673,
      "num_tokens": 9631019.0,
      "step": 5550
    },
    {
      "entropy": 5.740459442138672,
      "epoch": 0.43221163197821433,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985854245439408,
      "loss": 5.5368,
      "mean_token_accuracy": 0.15551533550024033,
      "num_tokens": 9639797.0,
      "step": 5555
    },
    {
      "entropy": 5.729131412506104,
      "epoch": 0.4326006613499319,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985823198802827,
      "loss": 5.5268,
      "mean_token_accuracy": 0.16611688137054442,
      "num_tokens": 9648393.0,
      "step": 5560
    },
    {
      "entropy": 5.744918298721314,
      "epoch": 0.4329896907216495,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985792118241188,
      "loss": 5.5115,
      "mean_token_accuracy": 0.16605060398578644,
      "num_tokens": 9657260.0,
      "step": 5565
    },
    {
      "entropy": 5.68444242477417,
      "epoch": 0.43337872009336703,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985761003754961,
      "loss": 5.5724,
      "mean_token_accuracy": 0.16387458592653276,
      "num_tokens": 9665883.0,
      "step": 5570
    },
    {
      "entropy": 5.801157808303833,
      "epoch": 0.4337677494650846,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985729855344622,
      "loss": 5.5804,
      "mean_token_accuracy": 0.16304195076227188,
      "num_tokens": 9674100.0,
      "step": 5575
    },
    {
      "entropy": 5.74932165145874,
      "epoch": 0.4341567788368022,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498569867301064,
      "loss": 5.5596,
      "mean_token_accuracy": 0.16284748315811157,
      "num_tokens": 9683236.0,
      "step": 5580
    },
    {
      "entropy": 5.716816759109497,
      "epoch": 0.43454580820851973,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985667456753489,
      "loss": 5.5241,
      "mean_token_accuracy": 0.16412055790424346,
      "num_tokens": 9692076.0,
      "step": 5585
    },
    {
      "entropy": 5.783638858795166,
      "epoch": 0.4349348375802373,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985636206573642,
      "loss": 5.5668,
      "mean_token_accuracy": 0.1635881170630455,
      "num_tokens": 9700266.0,
      "step": 5590
    },
    {
      "entropy": 5.719404172897339,
      "epoch": 0.43532386695195485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985604922471575,
      "loss": 5.5093,
      "mean_token_accuracy": 0.16096031665802002,
      "num_tokens": 9708643.0,
      "step": 5595
    },
    {
      "entropy": 5.772911167144775,
      "epoch": 0.43571289632367244,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985573604447761,
      "loss": 5.5724,
      "mean_token_accuracy": 0.15739189833402634,
      "num_tokens": 9717459.0,
      "step": 5600
    },
    {
      "entropy": 5.7623131275177,
      "epoch": 0.43610192569539,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985542252502676,
      "loss": 5.4626,
      "mean_token_accuracy": 0.17172451615333556,
      "num_tokens": 9726251.0,
      "step": 5605
    },
    {
      "entropy": 5.622464990615844,
      "epoch": 0.43649095506710756,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985510866636796,
      "loss": 5.4687,
      "mean_token_accuracy": 0.16263914555311204,
      "num_tokens": 9735327.0,
      "step": 5610
    },
    {
      "entropy": 5.7184959888458256,
      "epoch": 0.43687998443882514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985479446850596,
      "loss": 5.6235,
      "mean_token_accuracy": 0.15505748838186265,
      "num_tokens": 9743642.0,
      "step": 5615
    },
    {
      "entropy": 5.8103550434112545,
      "epoch": 0.4372690138105427,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985447993144554,
      "loss": 5.4808,
      "mean_token_accuracy": 0.1669997215270996,
      "num_tokens": 9752235.0,
      "step": 5620
    },
    {
      "entropy": 5.779250192642212,
      "epoch": 0.43765804318226026,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985416505519147,
      "loss": 5.5645,
      "mean_token_accuracy": 0.15476715564727783,
      "num_tokens": 9761012.0,
      "step": 5625
    },
    {
      "entropy": 5.729055547714234,
      "epoch": 0.43804707255397785,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985384983974853,
      "loss": 5.5337,
      "mean_token_accuracy": 0.16533123850822448,
      "num_tokens": 9769810.0,
      "step": 5630
    },
    {
      "entropy": 5.742403411865235,
      "epoch": 0.4384361019256954,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985353428512148,
      "loss": 5.4654,
      "mean_token_accuracy": 0.16260131299495698,
      "num_tokens": 9778186.0,
      "step": 5635
    },
    {
      "entropy": 5.691243028640747,
      "epoch": 0.43882513129741296,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985321839131514,
      "loss": 5.5343,
      "mean_token_accuracy": 0.16799721270799636,
      "num_tokens": 9786456.0,
      "step": 5640
    },
    {
      "entropy": 5.7298742771148685,
      "epoch": 0.4392141606691305,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985290215833428,
      "loss": 5.5325,
      "mean_token_accuracy": 0.15999200642108918,
      "num_tokens": 9794613.0,
      "step": 5645
    },
    {
      "entropy": 5.826755428314209,
      "epoch": 0.4396031900408481,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985258558618372,
      "loss": 5.4602,
      "mean_token_accuracy": 0.1681143507361412,
      "num_tokens": 9803388.0,
      "step": 5650
    },
    {
      "entropy": 5.708745288848877,
      "epoch": 0.43999221941256567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985226867486825,
      "loss": 5.5219,
      "mean_token_accuracy": 0.16547514200210572,
      "num_tokens": 9812450.0,
      "step": 5655
    },
    {
      "entropy": 5.727104377746582,
      "epoch": 0.4403812487842832,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985195142439267,
      "loss": 5.6087,
      "mean_token_accuracy": 0.15687978863716126,
      "num_tokens": 9821377.0,
      "step": 5660
    },
    {
      "entropy": 5.785880517959595,
      "epoch": 0.4407702781560008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985163383476181,
      "loss": 5.5096,
      "mean_token_accuracy": 0.1567897081375122,
      "num_tokens": 9829761.0,
      "step": 5665
    },
    {
      "entropy": 5.791961812973023,
      "epoch": 0.44115930752771837,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985131590598048,
      "loss": 5.5826,
      "mean_token_accuracy": 0.15582849308848382,
      "num_tokens": 9838921.0,
      "step": 5670
    },
    {
      "entropy": 5.67721734046936,
      "epoch": 0.4415483368994359,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985099763805352,
      "loss": 5.4196,
      "mean_token_accuracy": 0.17157016843557357,
      "num_tokens": 9846822.0,
      "step": 5675
    },
    {
      "entropy": 5.779719400405884,
      "epoch": 0.4419373662711535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985067903098574,
      "loss": 5.6384,
      "mean_token_accuracy": 0.16318747252225876,
      "num_tokens": 9855951.0,
      "step": 5680
    },
    {
      "entropy": 5.791326093673706,
      "epoch": 0.442326395642871,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985036008478197,
      "loss": 5.5891,
      "mean_token_accuracy": 0.15986811220645905,
      "num_tokens": 9864875.0,
      "step": 5685
    },
    {
      "entropy": 5.7927148818969725,
      "epoch": 0.4427154250145886,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985004079944707,
      "loss": 5.5382,
      "mean_token_accuracy": 0.16213968843221666,
      "num_tokens": 9873785.0,
      "step": 5690
    },
    {
      "entropy": 5.719140195846558,
      "epoch": 0.4431044543863062,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984972117498587,
      "loss": 5.5312,
      "mean_token_accuracy": 0.1606796622276306,
      "num_tokens": 9883214.0,
      "step": 5695
    },
    {
      "entropy": 5.683859729766846,
      "epoch": 0.4434934837580237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984940121140323,
      "loss": 5.5046,
      "mean_token_accuracy": 0.16533416211605073,
      "num_tokens": 9891346.0,
      "step": 5700
    },
    {
      "entropy": 5.694504690170288,
      "epoch": 0.4438825131297413,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049849080908704,
      "loss": 5.4428,
      "mean_token_accuracy": 0.16452051997184752,
      "num_tokens": 9900646.0,
      "step": 5705
    },
    {
      "entropy": 5.789317941665649,
      "epoch": 0.44427154250145884,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984876026689302,
      "loss": 5.5749,
      "mean_token_accuracy": 0.16129888445138932,
      "num_tokens": 9910851.0,
      "step": 5710
    },
    {
      "entropy": 5.792222166061402,
      "epoch": 0.4446605718731764,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498484392859752,
      "loss": 5.59,
      "mean_token_accuracy": 0.16029220074415207,
      "num_tokens": 9919919.0,
      "step": 5715
    },
    {
      "entropy": 5.815969800949096,
      "epoch": 0.445049601244894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984811796595538,
      "loss": 5.5239,
      "mean_token_accuracy": 0.15867404490709305,
      "num_tokens": 9928831.0,
      "step": 5720
    },
    {
      "entropy": 5.737311124801636,
      "epoch": 0.44543863061661154,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984779630683843,
      "loss": 5.48,
      "mean_token_accuracy": 0.16831743866205215,
      "num_tokens": 9936840.0,
      "step": 5725
    },
    {
      "entropy": 5.7112761497497555,
      "epoch": 0.44582765998832913,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984747430862924,
      "loss": 5.5599,
      "mean_token_accuracy": 0.16760646551847458,
      "num_tokens": 9945845.0,
      "step": 5730
    },
    {
      "entropy": 5.771790504455566,
      "epoch": 0.44621668936004666,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984715197133271,
      "loss": 5.5443,
      "mean_token_accuracy": 0.16156574189662934,
      "num_tokens": 9955176.0,
      "step": 5735
    },
    {
      "entropy": 5.661108493804932,
      "epoch": 0.44660571873176425,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984682929495371,
      "loss": 5.4356,
      "mean_token_accuracy": 0.17045264989137648,
      "num_tokens": 9963182.0,
      "step": 5740
    },
    {
      "entropy": 5.74065465927124,
      "epoch": 0.44699474810348183,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984650627949715,
      "loss": 5.5687,
      "mean_token_accuracy": 0.16232636719942092,
      "num_tokens": 9971239.0,
      "step": 5745
    },
    {
      "entropy": 5.7525311470031735,
      "epoch": 0.44738377747519936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984618292496792,
      "loss": 5.5176,
      "mean_token_accuracy": 0.1640138268470764,
      "num_tokens": 9980285.0,
      "step": 5750
    },
    {
      "entropy": 5.766064834594727,
      "epoch": 0.44777280684691695,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984585923137093,
      "loss": 5.558,
      "mean_token_accuracy": 0.1661367669701576,
      "num_tokens": 9988922.0,
      "step": 5755
    },
    {
      "entropy": 5.783611202239991,
      "epoch": 0.4481618362186345,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498455351987111,
      "loss": 5.6251,
      "mean_token_accuracy": 0.157930363714695,
      "num_tokens": 9998538.0,
      "step": 5760
    },
    {
      "entropy": 5.899012470245362,
      "epoch": 0.44855086559035207,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984521082699333,
      "loss": 5.5106,
      "mean_token_accuracy": 0.16188331991434096,
      "num_tokens": 10006646.0,
      "step": 5765
    },
    {
      "entropy": 5.7212669372558596,
      "epoch": 0.44893989496206965,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984488611622256,
      "loss": 5.5051,
      "mean_token_accuracy": 0.16526260673999787,
      "num_tokens": 10016511.0,
      "step": 5770
    },
    {
      "entropy": 5.652274227142334,
      "epoch": 0.4493289243337872,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498445610664037,
      "loss": 5.4141,
      "mean_token_accuracy": 0.16629135608673096,
      "num_tokens": 10025223.0,
      "step": 5775
    },
    {
      "entropy": 5.693808221817017,
      "epoch": 0.44971795370550477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498442356775417,
      "loss": 5.5032,
      "mean_token_accuracy": 0.1613439738750458,
      "num_tokens": 10033678.0,
      "step": 5780
    },
    {
      "entropy": 5.690686988830566,
      "epoch": 0.45010698307722236,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984390994964148,
      "loss": 5.4287,
      "mean_token_accuracy": 0.16981953531503677,
      "num_tokens": 10042028.0,
      "step": 5785
    },
    {
      "entropy": 5.72962589263916,
      "epoch": 0.4504960124489399,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049843583882708,
      "loss": 5.5768,
      "mean_token_accuracy": 0.1632343500852585,
      "num_tokens": 10050758.0,
      "step": 5790
    },
    {
      "entropy": 5.732880020141602,
      "epoch": 0.4508850418206575,
      "grad_norm": 1.0,
      "learning_rate": 0.000498432574767462,
      "loss": 5.3525,
      "mean_token_accuracy": 0.17709842324256897,
      "num_tokens": 10058688.0,
      "step": 5795
    },
    {
      "entropy": 5.650451421737671,
      "epoch": 0.451274071192375,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984293073176103,
      "loss": 5.4439,
      "mean_token_accuracy": 0.1742873877286911,
      "num_tokens": 10067302.0,
      "step": 5800
    },
    {
      "entropy": 5.644633388519287,
      "epoch": 0.4516631005640926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984260364775744,
      "loss": 5.4444,
      "mean_token_accuracy": 0.16964530050754548,
      "num_tokens": 10076424.0,
      "step": 5805
    },
    {
      "entropy": 5.7676972389221195,
      "epoch": 0.4520521299358102,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498422762247404,
      "loss": 5.571,
      "mean_token_accuracy": 0.16114722490310668,
      "num_tokens": 10086307.0,
      "step": 5810
    },
    {
      "entropy": 5.6775463104248045,
      "epoch": 0.4524411593075277,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984194846271489,
      "loss": 5.432,
      "mean_token_accuracy": 0.16395897418260574,
      "num_tokens": 10094954.0,
      "step": 5815
    },
    {
      "entropy": 5.702933311462402,
      "epoch": 0.4528301886792453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984162036168586,
      "loss": 5.5349,
      "mean_token_accuracy": 0.15665901601314544,
      "num_tokens": 10104401.0,
      "step": 5820
    },
    {
      "entropy": 5.675783395767212,
      "epoch": 0.4532192180509628,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984129192165831,
      "loss": 5.5369,
      "mean_token_accuracy": 0.1650775119662285,
      "num_tokens": 10112596.0,
      "step": 5825
    },
    {
      "entropy": 5.718826484680176,
      "epoch": 0.4536082474226804,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984096314263722,
      "loss": 5.5792,
      "mean_token_accuracy": 0.16326392143964769,
      "num_tokens": 10121311.0,
      "step": 5830
    },
    {
      "entropy": 5.678861618041992,
      "epoch": 0.453997276794398,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984063402462757,
      "loss": 5.3773,
      "mean_token_accuracy": 0.17142716199159622,
      "num_tokens": 10130010.0,
      "step": 5835
    },
    {
      "entropy": 5.707784795761109,
      "epoch": 0.45438630616611553,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984030456763435,
      "loss": 5.4764,
      "mean_token_accuracy": 0.1730629026889801,
      "num_tokens": 10138473.0,
      "step": 5840
    },
    {
      "entropy": 5.8273272037506105,
      "epoch": 0.4547753355378331,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983997477166257,
      "loss": 5.5745,
      "mean_token_accuracy": 0.16354735046625138,
      "num_tokens": 10146843.0,
      "step": 5845
    },
    {
      "entropy": 5.625149297714233,
      "epoch": 0.45516436490955064,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983964463671723,
      "loss": 5.4182,
      "mean_token_accuracy": 0.16891502290964128,
      "num_tokens": 10155319.0,
      "step": 5850
    },
    {
      "entropy": 5.696143960952758,
      "epoch": 0.45555339428126823,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983931416280334,
      "loss": 5.6027,
      "mean_token_accuracy": 0.15458251535892487,
      "num_tokens": 10164308.0,
      "step": 5855
    },
    {
      "entropy": 5.793465566635132,
      "epoch": 0.4559424236529858,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983898334992591,
      "loss": 5.6094,
      "mean_token_accuracy": 0.1615153282880783,
      "num_tokens": 10173173.0,
      "step": 5860
    },
    {
      "entropy": 5.764628934860229,
      "epoch": 0.45633145302470335,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983865219808998,
      "loss": 5.5159,
      "mean_token_accuracy": 0.16485823392868043,
      "num_tokens": 10181376.0,
      "step": 5865
    },
    {
      "entropy": 5.81520094871521,
      "epoch": 0.45672048239642093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983832070730055,
      "loss": 5.6275,
      "mean_token_accuracy": 0.15767347514629365,
      "num_tokens": 10189701.0,
      "step": 5870
    },
    {
      "entropy": 5.751556491851806,
      "epoch": 0.4571095117681385,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983798887756265,
      "loss": 5.4879,
      "mean_token_accuracy": 0.1661074161529541,
      "num_tokens": 10198721.0,
      "step": 5875
    },
    {
      "entropy": 5.677633237838745,
      "epoch": 0.45749854113985605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983765670888133,
      "loss": 5.3744,
      "mean_token_accuracy": 0.16815058141946793,
      "num_tokens": 10206428.0,
      "step": 5880
    },
    {
      "entropy": 5.633647203445435,
      "epoch": 0.45788757051157364,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983732420126163,
      "loss": 5.5197,
      "mean_token_accuracy": 0.16477368921041488,
      "num_tokens": 10214941.0,
      "step": 5885
    },
    {
      "entropy": 5.7348369598388675,
      "epoch": 0.45827659988329117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983699135470858,
      "loss": 5.5078,
      "mean_token_accuracy": 0.16581415385007858,
      "num_tokens": 10223491.0,
      "step": 5890
    },
    {
      "entropy": 5.737116384506225,
      "epoch": 0.45866562925500876,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983665816922723,
      "loss": 5.4472,
      "mean_token_accuracy": 0.17093681842088698,
      "num_tokens": 10231820.0,
      "step": 5895
    },
    {
      "entropy": 5.6952796459198,
      "epoch": 0.45905465862672634,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983632464482267,
      "loss": 5.4493,
      "mean_token_accuracy": 0.16853009462356566,
      "num_tokens": 10239781.0,
      "step": 5900
    },
    {
      "entropy": 5.739758110046386,
      "epoch": 0.4594436879984439,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983599078149991,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17133889049291612,
      "num_tokens": 10248093.0,
      "step": 5905
    },
    {
      "entropy": 5.763691329956055,
      "epoch": 0.45983271737016146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983565657926405,
      "loss": 5.5025,
      "mean_token_accuracy": 0.16365934312343597,
      "num_tokens": 10256370.0,
      "step": 5910
    },
    {
      "entropy": 5.643654251098633,
      "epoch": 0.460221746741879,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983532203812017,
      "loss": 5.4761,
      "mean_token_accuracy": 0.16236139535903932,
      "num_tokens": 10264924.0,
      "step": 5915
    },
    {
      "entropy": 5.700886058807373,
      "epoch": 0.4606107761135966,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983498715807331,
      "loss": 5.4139,
      "mean_token_accuracy": 0.1686012089252472,
      "num_tokens": 10273488.0,
      "step": 5920
    },
    {
      "entropy": 5.712774229049683,
      "epoch": 0.46099980548531416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983465193912857,
      "loss": 5.4892,
      "mean_token_accuracy": 0.16322697699069977,
      "num_tokens": 10281534.0,
      "step": 5925
    },
    {
      "entropy": 5.709554719924927,
      "epoch": 0.4613888348570317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983431638129104,
      "loss": 5.5247,
      "mean_token_accuracy": 0.1624474197626114,
      "num_tokens": 10289613.0,
      "step": 5930
    },
    {
      "entropy": 5.7896459102630615,
      "epoch": 0.4617778642287493,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983398048456581,
      "loss": 5.5811,
      "mean_token_accuracy": 0.16493575274944305,
      "num_tokens": 10299027.0,
      "step": 5935
    },
    {
      "entropy": 5.764388799667358,
      "epoch": 0.4621668936004668,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983364424895796,
      "loss": 5.5387,
      "mean_token_accuracy": 0.16037665605545043,
      "num_tokens": 10308727.0,
      "step": 5940
    },
    {
      "entropy": 5.74548830986023,
      "epoch": 0.4625559229721844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983330767447262,
      "loss": 5.5594,
      "mean_token_accuracy": 0.16487368792295456,
      "num_tokens": 10318129.0,
      "step": 5945
    },
    {
      "entropy": 5.677147579193115,
      "epoch": 0.462944952343902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983297076111489,
      "loss": 5.3962,
      "mean_token_accuracy": 0.1714185506105423,
      "num_tokens": 10326504.0,
      "step": 5950
    },
    {
      "entropy": 5.625962829589843,
      "epoch": 0.4633339817156195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983263350888987,
      "loss": 5.4669,
      "mean_token_accuracy": 0.16761284768581391,
      "num_tokens": 10334585.0,
      "step": 5955
    },
    {
      "entropy": 5.667236328125,
      "epoch": 0.4637230110873371,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983229591780268,
      "loss": 5.4601,
      "mean_token_accuracy": 0.17346582263708116,
      "num_tokens": 10343328.0,
      "step": 5960
    },
    {
      "entropy": 5.694534826278686,
      "epoch": 0.46411204045905463,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983195798785844,
      "loss": 5.5072,
      "mean_token_accuracy": 0.16164835244417192,
      "num_tokens": 10352289.0,
      "step": 5965
    },
    {
      "entropy": 5.788618087768555,
      "epoch": 0.4645010698307722,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983161971906228,
      "loss": 5.6119,
      "mean_token_accuracy": 0.155012309551239,
      "num_tokens": 10360847.0,
      "step": 5970
    },
    {
      "entropy": 5.755517053604126,
      "epoch": 0.4648900992024898,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983128111141935,
      "loss": 5.5251,
      "mean_token_accuracy": 0.1676684468984604,
      "num_tokens": 10368911.0,
      "step": 5975
    },
    {
      "entropy": 5.706905555725098,
      "epoch": 0.46527912857420733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983094216493475,
      "loss": 5.4706,
      "mean_token_accuracy": 0.1699859157204628,
      "num_tokens": 10377257.0,
      "step": 5980
    },
    {
      "entropy": 5.702120542526245,
      "epoch": 0.4656681579459249,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983060287961367,
      "loss": 5.5287,
      "mean_token_accuracy": 0.16288657635450363,
      "num_tokens": 10386167.0,
      "step": 5985
    },
    {
      "entropy": 5.679756593704224,
      "epoch": 0.4660571873176425,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983026325546123,
      "loss": 5.4184,
      "mean_token_accuracy": 0.16258866041898729,
      "num_tokens": 10395802.0,
      "step": 5990
    },
    {
      "entropy": 5.699802398681641,
      "epoch": 0.46644621668936004,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982992329248257,
      "loss": 5.5527,
      "mean_token_accuracy": 0.1626785546541214,
      "num_tokens": 10404836.0,
      "step": 5995
    },
    {
      "entropy": 5.770925092697143,
      "epoch": 0.4668352460610776,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982958299068289,
      "loss": 5.5062,
      "mean_token_accuracy": 0.16537801325321197,
      "num_tokens": 10413394.0,
      "step": 6000
    },
    {
      "epoch": 0.4668352460610776,
      "eval_entropy": 5.524670014367641,
      "eval_loss": 5.532135963439941,
      "eval_mean_token_accuracy": 0.17199271473463507,
      "eval_num_tokens": 10413394.0,
      "eval_runtime": 21.8497,
      "eval_samples_per_second": 1901.998,
      "eval_steps_per_second": 237.761,
      "step": 6000
    }
  ],
  "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": 2095486464983040.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}