{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0596962204168139,
  "eval_steps": 3000,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.937930107116699,
      "epoch": 0.0004415400918403391,
      "grad_norm": 16.25,
      "learning_rate": 2e-06,
      "loss": 14.1689,
      "mean_token_accuracy": 0.0,
      "num_tokens": 9295.0,
      "step": 5
    },
    {
      "entropy": 4.9478801727294925,
      "epoch": 0.0008830801836806782,
      "grad_norm": 16.75,
      "learning_rate": 4.5e-06,
      "loss": 14.0804,
      "mean_token_accuracy": 0.0,
      "num_tokens": 17694.0,
      "step": 10
    },
    {
      "entropy": 4.970736074447632,
      "epoch": 0.0013246202755210173,
      "grad_norm": 17.875,
      "learning_rate": 7e-06,
      "loss": 14.0763,
      "mean_token_accuracy": 0.0,
      "num_tokens": 26875.0,
      "step": 15
    },
    {
      "entropy": 5.00352029800415,
      "epoch": 0.0017661603673613563,
      "grad_norm": 20.625,
      "learning_rate": 9.5e-06,
      "loss": 13.9684,
      "mean_token_accuracy": 9.208103292621673e-05,
      "num_tokens": 38072.0,
      "step": 20
    },
    {
      "entropy": 5.193281745910644,
      "epoch": 0.0022077004592016957,
      "grad_norm": 25.875,
      "learning_rate": 1.2e-05,
      "loss": 13.4895,
      "mean_token_accuracy": 0.00010976948542520403,
      "num_tokens": 48899.0,
      "step": 25
    },
    {
      "entropy": 5.656161642074585,
      "epoch": 0.0026492405510420347,
      "grad_norm": 31.0,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 12.7498,
      "mean_token_accuracy": 0.0,
      "num_tokens": 57994.0,
      "step": 30
    },
    {
      "entropy": 7.820593309402466,
      "epoch": 0.0030907806428823736,
      "grad_norm": 14.5625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.581,
      "mean_token_accuracy": 0.0,
      "num_tokens": 67252.0,
      "step": 35
    },
    {
      "entropy": 10.46483039855957,
      "epoch": 0.0035323207347227126,
      "grad_norm": 3.5625,
      "learning_rate": 1.95e-05,
      "loss": 10.7728,
      "mean_token_accuracy": 0.00608623722800985,
      "num_tokens": 76787.0,
      "step": 40
    },
    {
      "entropy": 10.734039402008056,
      "epoch": 0.003973860826563052,
      "grad_norm": 3.15625,
      "learning_rate": 2.2e-05,
      "loss": 10.6475,
      "mean_token_accuracy": 0.016124986205250026,
      "num_tokens": 86718.0,
      "step": 45
    },
    {
      "entropy": 10.734684371948243,
      "epoch": 0.004415400918403391,
      "grad_norm": 2.96875,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.444,
      "mean_token_accuracy": 0.017893549520522356,
      "num_tokens": 96136.0,
      "step": 50
    },
    {
      "entropy": 10.690448665618897,
      "epoch": 0.00485694101024373,
      "grad_norm": 2.40625,
      "learning_rate": 2.7e-05,
      "loss": 10.2836,
      "mean_token_accuracy": 0.016107181273400783,
      "num_tokens": 106579.0,
      "step": 55
    },
    {
      "entropy": 10.6747314453125,
      "epoch": 0.005298481102084069,
      "grad_norm": 2.296875,
      "learning_rate": 2.95e-05,
      "loss": 10.0794,
      "mean_token_accuracy": 0.01910884603857994,
      "num_tokens": 115698.0,
      "step": 60
    },
    {
      "entropy": 10.712893676757812,
      "epoch": 0.005740021193924408,
      "grad_norm": 1.796875,
      "learning_rate": 3.2e-05,
      "loss": 9.987,
      "mean_token_accuracy": 0.04295358825474978,
      "num_tokens": 125141.0,
      "step": 65
    },
    {
      "entropy": 10.70653076171875,
      "epoch": 0.006181561285764747,
      "grad_norm": 1.921875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8323,
      "mean_token_accuracy": 0.06280341930687428,
      "num_tokens": 134148.0,
      "step": 70
    },
    {
      "entropy": 10.686304759979247,
      "epoch": 0.006623101377605086,
      "grad_norm": 1.7421875,
      "learning_rate": 3.7e-05,
      "loss": 9.8022,
      "mean_token_accuracy": 0.05989494659006596,
      "num_tokens": 143032.0,
      "step": 75
    },
    {
      "entropy": 10.672717380523682,
      "epoch": 0.007064641469445425,
      "grad_norm": 1.78125,
      "learning_rate": 3.95e-05,
      "loss": 9.7113,
      "mean_token_accuracy": 0.06549292616546154,
      "num_tokens": 151477.0,
      "step": 80
    },
    {
      "entropy": 10.66338300704956,
      "epoch": 0.007506181561285765,
      "grad_norm": 1.8984375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6711,
      "mean_token_accuracy": 0.06344555392861366,
      "num_tokens": 161364.0,
      "step": 85
    },
    {
      "entropy": 10.649686336517334,
      "epoch": 0.007947721653126103,
      "grad_norm": 1.875,
      "learning_rate": 4.45e-05,
      "loss": 9.5988,
      "mean_token_accuracy": 0.059225669503211974,
      "num_tokens": 171771.0,
      "step": 90
    },
    {
      "entropy": 10.630746650695801,
      "epoch": 0.008389261744966443,
      "grad_norm": 1.9296875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4381,
      "mean_token_accuracy": 0.06350292377173901,
      "num_tokens": 180461.0,
      "step": 95
    },
    {
      "entropy": 10.59743824005127,
      "epoch": 0.008830801836806783,
      "grad_norm": 1.703125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4588,
      "mean_token_accuracy": 0.06274225264787674,
      "num_tokens": 190873.0,
      "step": 100
    },
    {
      "entropy": 10.572507667541505,
      "epoch": 0.009272341928647121,
      "grad_norm": 1.6796875,
      "learning_rate": 5.2e-05,
      "loss": 9.2659,
      "mean_token_accuracy": 0.06957426071166992,
      "num_tokens": 199146.0,
      "step": 105
    },
    {
      "entropy": 10.532439994812012,
      "epoch": 0.00971388202048746,
      "grad_norm": 1.671875,
      "learning_rate": 5.45e-05,
      "loss": 9.2169,
      "mean_token_accuracy": 0.07027090489864349,
      "num_tokens": 208218.0,
      "step": 110
    },
    {
      "entropy": 10.482099723815917,
      "epoch": 0.010155422112327799,
      "grad_norm": 1.6171875,
      "learning_rate": 5.7e-05,
      "loss": 9.1174,
      "mean_token_accuracy": 0.0710879024118185,
      "num_tokens": 217548.0,
      "step": 115
    },
    {
      "entropy": 10.431210136413574,
      "epoch": 0.010596962204168139,
      "grad_norm": 1.53125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 9.0402,
      "mean_token_accuracy": 0.06714637242257596,
      "num_tokens": 226157.0,
      "step": 120
    },
    {
      "entropy": 10.364525890350341,
      "epoch": 0.011038502296008477,
      "grad_norm": 1.359375,
      "learning_rate": 6.2e-05,
      "loss": 8.957,
      "mean_token_accuracy": 0.07038320675492286,
      "num_tokens": 235879.0,
      "step": 125
    },
    {
      "entropy": 10.272084522247315,
      "epoch": 0.011480042387848817,
      "grad_norm": 1.40625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.8401,
      "mean_token_accuracy": 0.07076714783906937,
      "num_tokens": 244532.0,
      "step": 130
    },
    {
      "entropy": 10.150115776062012,
      "epoch": 0.011921582479689156,
      "grad_norm": 1.3515625,
      "learning_rate": 6.7e-05,
      "loss": 8.7849,
      "mean_token_accuracy": 0.07159631326794624,
      "num_tokens": 254037.0,
      "step": 135
    },
    {
      "entropy": 10.064141464233398,
      "epoch": 0.012363122571529495,
      "grad_norm": 1.28125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.7124,
      "mean_token_accuracy": 0.06714138388633728,
      "num_tokens": 264162.0,
      "step": 140
    },
    {
      "entropy": 9.97827501296997,
      "epoch": 0.012804662663369834,
      "grad_norm": 1.359375,
      "learning_rate": 7.2e-05,
      "loss": 8.6055,
      "mean_token_accuracy": 0.07109769769012927,
      "num_tokens": 273644.0,
      "step": 145
    },
    {
      "entropy": 9.83043041229248,
      "epoch": 0.013246202755210172,
      "grad_norm": 1.0390625,
      "learning_rate": 7.45e-05,
      "loss": 8.4902,
      "mean_token_accuracy": 0.07540867626667022,
      "num_tokens": 283643.0,
      "step": 150
    },
    {
      "entropy": 9.671627426147461,
      "epoch": 0.013687742847050512,
      "grad_norm": 1.4375,
      "learning_rate": 7.7e-05,
      "loss": 8.3939,
      "mean_token_accuracy": 0.0741175465285778,
      "num_tokens": 292636.0,
      "step": 155
    },
    {
      "entropy": 9.572625064849854,
      "epoch": 0.01412928293889085,
      "grad_norm": 1.0546875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.2828,
      "mean_token_accuracy": 0.07567049115896225,
      "num_tokens": 302059.0,
      "step": 160
    },
    {
      "entropy": 9.38239164352417,
      "epoch": 0.01457082303073119,
      "grad_norm": 1.03125,
      "learning_rate": 8.2e-05,
      "loss": 8.3316,
      "mean_token_accuracy": 0.07153738178312778,
      "num_tokens": 311995.0,
      "step": 165
    },
    {
      "entropy": 9.255543613433838,
      "epoch": 0.01501236312257153,
      "grad_norm": 0.9375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.2309,
      "mean_token_accuracy": 0.07089664451777936,
      "num_tokens": 322035.0,
      "step": 170
    },
    {
      "entropy": 9.142601108551025,
      "epoch": 0.015453903214411868,
      "grad_norm": 0.9765625,
      "learning_rate": 8.7e-05,
      "loss": 8.1473,
      "mean_token_accuracy": 0.07471434287726879,
      "num_tokens": 330746.0,
      "step": 175
    },
    {
      "entropy": 8.977528285980224,
      "epoch": 0.015895443306252206,
      "grad_norm": 1.0703125,
      "learning_rate": 8.95e-05,
      "loss": 8.1055,
      "mean_token_accuracy": 0.07090946286916733,
      "num_tokens": 339445.0,
      "step": 180
    },
    {
      "entropy": 8.824779510498047,
      "epoch": 0.016336983398092548,
      "grad_norm": 0.76171875,
      "learning_rate": 9.2e-05,
      "loss": 8.1189,
      "mean_token_accuracy": 0.06805617697536945,
      "num_tokens": 349171.0,
      "step": 185
    },
    {
      "entropy": 8.705509662628174,
      "epoch": 0.016778523489932886,
      "grad_norm": 0.68359375,
      "learning_rate": 9.45e-05,
      "loss": 8.0342,
      "mean_token_accuracy": 0.07489757090806962,
      "num_tokens": 358146.0,
      "step": 190
    },
    {
      "entropy": 8.5638991355896,
      "epoch": 0.017220063581773224,
      "grad_norm": 0.7421875,
      "learning_rate": 9.7e-05,
      "loss": 7.9875,
      "mean_token_accuracy": 0.07784983068704605,
      "num_tokens": 366921.0,
      "step": 195
    },
    {
      "entropy": 8.432452964782716,
      "epoch": 0.017661603673613566,
      "grad_norm": 0.7109375,
      "learning_rate": 9.95e-05,
      "loss": 8.01,
      "mean_token_accuracy": 0.07490835711359978,
      "num_tokens": 375993.0,
      "step": 200
    },
    {
      "entropy": 8.46620798110962,
      "epoch": 0.018103143765453904,
      "grad_norm": 0.67578125,
      "learning_rate": 0.000102,
      "loss": 8.1292,
      "mean_token_accuracy": 0.06469435021281242,
      "num_tokens": 385072.0,
      "step": 205
    },
    {
      "entropy": 8.469141292572022,
      "epoch": 0.018544683857294242,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.9661,
      "mean_token_accuracy": 0.07081307880580426,
      "num_tokens": 393928.0,
      "step": 210
    },
    {
      "entropy": 8.398229122161865,
      "epoch": 0.01898622394913458,
      "grad_norm": 0.64453125,
      "learning_rate": 0.000107,
      "loss": 8.0343,
      "mean_token_accuracy": 0.07047690935432911,
      "num_tokens": 403245.0,
      "step": 215
    },
    {
      "entropy": 8.3765549659729,
      "epoch": 0.01942776404097492,
      "grad_norm": 0.625,
      "learning_rate": 0.0001095,
      "loss": 7.9137,
      "mean_token_accuracy": 0.08345218449831009,
      "num_tokens": 413192.0,
      "step": 220
    },
    {
      "entropy": 8.40019121170044,
      "epoch": 0.01986930413281526,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000112,
      "loss": 7.9652,
      "mean_token_accuracy": 0.07346240207552909,
      "num_tokens": 422219.0,
      "step": 225
    },
    {
      "entropy": 8.349788379669189,
      "epoch": 0.020310844224655598,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0001145,
      "loss": 7.9015,
      "mean_token_accuracy": 0.08022939898073674,
      "num_tokens": 431654.0,
      "step": 230
    },
    {
      "entropy": 8.380678462982178,
      "epoch": 0.02075238431649594,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.9143,
      "mean_token_accuracy": 0.08441097363829612,
      "num_tokens": 441817.0,
      "step": 235
    },
    {
      "entropy": 8.25525884628296,
      "epoch": 0.021193924408336277,
      "grad_norm": 0.796875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.8807,
      "mean_token_accuracy": 0.08318600244820118,
      "num_tokens": 450285.0,
      "step": 240
    },
    {
      "entropy": 8.265483093261718,
      "epoch": 0.021635464500176615,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000122,
      "loss": 7.9189,
      "mean_token_accuracy": 0.07491880282759666,
      "num_tokens": 459726.0,
      "step": 245
    },
    {
      "entropy": 8.335671520233154,
      "epoch": 0.022077004592016954,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0001245,
      "loss": 7.8952,
      "mean_token_accuracy": 0.07705423757433891,
      "num_tokens": 469236.0,
      "step": 250
    },
    {
      "entropy": 8.266505146026612,
      "epoch": 0.022518544683857295,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000127,
      "loss": 7.9973,
      "mean_token_accuracy": 0.07031807117164135,
      "num_tokens": 478835.0,
      "step": 255
    },
    {
      "entropy": 8.278965950012207,
      "epoch": 0.022960084775697633,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0001295,
      "loss": 7.8611,
      "mean_token_accuracy": 0.07837036736309529,
      "num_tokens": 487748.0,
      "step": 260
    },
    {
      "entropy": 8.250927925109863,
      "epoch": 0.02340162486753797,
      "grad_norm": 0.69140625,
      "learning_rate": 0.000132,
      "loss": 7.976,
      "mean_token_accuracy": 0.07374264337122441,
      "num_tokens": 497733.0,
      "step": 265
    },
    {
      "entropy": 8.279707717895509,
      "epoch": 0.023843164959378313,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.8168,
      "mean_token_accuracy": 0.0817179311066866,
      "num_tokens": 506351.0,
      "step": 270
    },
    {
      "entropy": 8.150735473632812,
      "epoch": 0.02428470505121865,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.8459,
      "mean_token_accuracy": 0.08556010201573372,
      "num_tokens": 515865.0,
      "step": 275
    },
    {
      "entropy": 8.268842601776123,
      "epoch": 0.02472624514305899,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001395,
      "loss": 7.8569,
      "mean_token_accuracy": 0.0806879721581936,
      "num_tokens": 524695.0,
      "step": 280
    },
    {
      "entropy": 8.203614711761475,
      "epoch": 0.025167785234899327,
      "grad_norm": 0.890625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.8531,
      "mean_token_accuracy": 0.07765759006142617,
      "num_tokens": 533896.0,
      "step": 285
    },
    {
      "entropy": 8.160596179962159,
      "epoch": 0.02560932532673967,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0001445,
      "loss": 7.9379,
      "mean_token_accuracy": 0.072020772844553,
      "num_tokens": 543260.0,
      "step": 290
    },
    {
      "entropy": 8.251058101654053,
      "epoch": 0.026050865418580007,
      "grad_norm": 0.7421875,
      "learning_rate": 0.000147,
      "loss": 7.8555,
      "mean_token_accuracy": 0.0810875877737999,
      "num_tokens": 553857.0,
      "step": 295
    },
    {
      "entropy": 8.141295623779296,
      "epoch": 0.026492405510420345,
      "grad_norm": 0.796875,
      "learning_rate": 0.0001495,
      "loss": 7.821,
      "mean_token_accuracy": 0.0825580857694149,
      "num_tokens": 562461.0,
      "step": 300
    },
    {
      "entropy": 8.181407356262207,
      "epoch": 0.026933945602260687,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000152,
      "loss": 7.8487,
      "mean_token_accuracy": 0.07985757477581501,
      "num_tokens": 571863.0,
      "step": 305
    },
    {
      "entropy": 8.115688610076905,
      "epoch": 0.027375485694101025,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.8567,
      "mean_token_accuracy": 0.07967338487505912,
      "num_tokens": 581063.0,
      "step": 310
    },
    {
      "entropy": 8.166765213012695,
      "epoch": 0.027817025785941363,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000157,
      "loss": 7.7804,
      "mean_token_accuracy": 0.08055768013000489,
      "num_tokens": 590821.0,
      "step": 315
    },
    {
      "entropy": 8.144050884246827,
      "epoch": 0.0282585658777817,
      "grad_norm": 0.96875,
      "learning_rate": 0.0001595,
      "loss": 7.754,
      "mean_token_accuracy": 0.0802922174334526,
      "num_tokens": 599701.0,
      "step": 320
    },
    {
      "entropy": 8.107472562789917,
      "epoch": 0.028700105969622042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000162,
      "loss": 7.7741,
      "mean_token_accuracy": 0.07862499617040157,
      "num_tokens": 608071.0,
      "step": 325
    },
    {
      "entropy": 8.11353440284729,
      "epoch": 0.02914164606146238,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.8263,
      "mean_token_accuracy": 0.07856386005878449,
      "num_tokens": 616643.0,
      "step": 330
    },
    {
      "entropy": 8.142715454101562,
      "epoch": 0.02958318615330272,
      "grad_norm": 1.0,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.8415,
      "mean_token_accuracy": 0.08129560798406602,
      "num_tokens": 626002.0,
      "step": 335
    },
    {
      "entropy": 8.08275752067566,
      "epoch": 0.03002472624514306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.7904,
      "mean_token_accuracy": 0.08131296187639236,
      "num_tokens": 636100.0,
      "step": 340
    },
    {
      "entropy": 8.157522010803223,
      "epoch": 0.0304662663369834,
      "grad_norm": 1.015625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.8174,
      "mean_token_accuracy": 0.07889350652694702,
      "num_tokens": 645142.0,
      "step": 345
    },
    {
      "entropy": 8.096787357330323,
      "epoch": 0.030907806428823736,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.7584,
      "mean_token_accuracy": 0.0830284185707569,
      "num_tokens": 654482.0,
      "step": 350
    },
    {
      "entropy": 8.052792024612426,
      "epoch": 0.031349346520664075,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000177,
      "loss": 7.7812,
      "mean_token_accuracy": 0.07971454188227653,
      "num_tokens": 663167.0,
      "step": 355
    },
    {
      "entropy": 8.085110521316528,
      "epoch": 0.03179088661250441,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0001795,
      "loss": 7.8482,
      "mean_token_accuracy": 0.0779881864786148,
      "num_tokens": 672996.0,
      "step": 360
    },
    {
      "entropy": 8.109179019927979,
      "epoch": 0.03223242670434476,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000182,
      "loss": 7.778,
      "mean_token_accuracy": 0.0821885697543621,
      "num_tokens": 681479.0,
      "step": 365
    },
    {
      "entropy": 8.096177530288696,
      "epoch": 0.032673966796185096,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0001845,
      "loss": 7.699,
      "mean_token_accuracy": 0.08528197333216667,
      "num_tokens": 690222.0,
      "step": 370
    },
    {
      "entropy": 8.048606014251709,
      "epoch": 0.033115506888025434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000187,
      "loss": 7.7136,
      "mean_token_accuracy": 0.08710152879357339,
      "num_tokens": 699143.0,
      "step": 375
    },
    {
      "entropy": 7.981799745559693,
      "epoch": 0.03355704697986577,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001895,
      "loss": 7.6844,
      "mean_token_accuracy": 0.08963175565004348,
      "num_tokens": 708416.0,
      "step": 380
    },
    {
      "entropy": 8.0686354637146,
      "epoch": 0.03399858707170611,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000192,
      "loss": 7.7816,
      "mean_token_accuracy": 0.08127610832452774,
      "num_tokens": 718140.0,
      "step": 385
    },
    {
      "entropy": 8.01650652885437,
      "epoch": 0.03444012716354645,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0001945,
      "loss": 7.7234,
      "mean_token_accuracy": 0.08251546621322632,
      "num_tokens": 727100.0,
      "step": 390
    },
    {
      "entropy": 7.99934663772583,
      "epoch": 0.034881667255386786,
      "grad_norm": 0.84375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.6959,
      "mean_token_accuracy": 0.0842419758439064,
      "num_tokens": 736383.0,
      "step": 395
    },
    {
      "entropy": 8.089943981170654,
      "epoch": 0.03532320734722713,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.755,
      "mean_token_accuracy": 0.08859959691762924,
      "num_tokens": 745746.0,
      "step": 400
    },
    {
      "entropy": 8.041479778289794,
      "epoch": 0.03576474743906747,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000202,
      "loss": 7.7302,
      "mean_token_accuracy": 0.08785694614052772,
      "num_tokens": 754276.0,
      "step": 405
    },
    {
      "entropy": 8.044493007659913,
      "epoch": 0.03620628753090781,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.6459,
      "mean_token_accuracy": 0.08876177445054054,
      "num_tokens": 763044.0,
      "step": 410
    },
    {
      "entropy": 8.00500979423523,
      "epoch": 0.036647827622748146,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000207,
      "loss": 7.6932,
      "mean_token_accuracy": 0.08604336977005005,
      "num_tokens": 771727.0,
      "step": 415
    },
    {
      "entropy": 8.017408561706542,
      "epoch": 0.037089367714588484,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002095,
      "loss": 7.6466,
      "mean_token_accuracy": 0.08596126586198807,
      "num_tokens": 781430.0,
      "step": 420
    },
    {
      "entropy": 7.938171768188477,
      "epoch": 0.03753090780642882,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000212,
      "loss": 7.6106,
      "mean_token_accuracy": 0.09112684056162834,
      "num_tokens": 790619.0,
      "step": 425
    },
    {
      "entropy": 8.108117580413818,
      "epoch": 0.03797244789826916,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002145,
      "loss": 7.7037,
      "mean_token_accuracy": 0.08285126835107803,
      "num_tokens": 800264.0,
      "step": 430
    },
    {
      "entropy": 8.017135334014892,
      "epoch": 0.038413987990109505,
      "grad_norm": 1.125,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.6698,
      "mean_token_accuracy": 0.08375026062130927,
      "num_tokens": 810522.0,
      "step": 435
    },
    {
      "entropy": 7.982311105728149,
      "epoch": 0.03885552808194984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002195,
      "loss": 7.6376,
      "mean_token_accuracy": 0.08961728811264039,
      "num_tokens": 819361.0,
      "step": 440
    },
    {
      "entropy": 7.917629718780518,
      "epoch": 0.03929706817379018,
      "grad_norm": 0.84375,
      "learning_rate": 0.000222,
      "loss": 7.681,
      "mean_token_accuracy": 0.08837661817669869,
      "num_tokens": 829146.0,
      "step": 445
    },
    {
      "entropy": 7.935456418991089,
      "epoch": 0.03973860826563052,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002245,
      "loss": 7.5834,
      "mean_token_accuracy": 0.09174824506044388,
      "num_tokens": 837315.0,
      "step": 450
    },
    {
      "entropy": 7.988494825363159,
      "epoch": 0.04018014835747086,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.599,
      "mean_token_accuracy": 0.08744515627622604,
      "num_tokens": 845757.0,
      "step": 455
    },
    {
      "entropy": 7.945766162872315,
      "epoch": 0.040621688449311195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.6764,
      "mean_token_accuracy": 0.08117738515138626,
      "num_tokens": 854899.0,
      "step": 460
    },
    {
      "entropy": 7.949225807189942,
      "epoch": 0.041063228541151534,
      "grad_norm": 2.046875,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.5731,
      "mean_token_accuracy": 0.09045649319887161,
      "num_tokens": 863839.0,
      "step": 465
    },
    {
      "entropy": 7.848576593399048,
      "epoch": 0.04150476863299188,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.4955,
      "mean_token_accuracy": 0.09183765053749085,
      "num_tokens": 872249.0,
      "step": 470
    },
    {
      "entropy": 7.883409404754639,
      "epoch": 0.04194630872483222,
      "grad_norm": 1.1875,
      "learning_rate": 0.000237,
      "loss": 7.6401,
      "mean_token_accuracy": 0.08584671169519424,
      "num_tokens": 881839.0,
      "step": 475
    },
    {
      "entropy": 8.046848773956299,
      "epoch": 0.042387848816672555,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002395,
      "loss": 7.66,
      "mean_token_accuracy": 0.08414728939533234,
      "num_tokens": 890865.0,
      "step": 480
    },
    {
      "entropy": 7.87333869934082,
      "epoch": 0.04282938890851289,
      "grad_norm": 1.046875,
      "learning_rate": 0.000242,
      "loss": 7.6421,
      "mean_token_accuracy": 0.08928802534937859,
      "num_tokens": 900858.0,
      "step": 485
    },
    {
      "entropy": 7.997522354125977,
      "epoch": 0.04327092900035323,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002445,
      "loss": 7.6155,
      "mean_token_accuracy": 0.08748602122068405,
      "num_tokens": 910396.0,
      "step": 490
    },
    {
      "entropy": 7.981202268600464,
      "epoch": 0.04371246909219357,
      "grad_norm": 1.0625,
      "learning_rate": 0.000247,
      "loss": 7.5676,
      "mean_token_accuracy": 0.09339846298098564,
      "num_tokens": 919532.0,
      "step": 495
    },
    {
      "entropy": 7.883893203735352,
      "epoch": 0.04415400918403391,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002495,
      "loss": 7.5701,
      "mean_token_accuracy": 0.09045918881893159,
      "num_tokens": 928204.0,
      "step": 500
    },
    {
      "entropy": 7.865892791748047,
      "epoch": 0.04459554927587425,
      "grad_norm": 1.25,
      "learning_rate": 0.000252,
      "loss": 7.5447,
      "mean_token_accuracy": 0.09523545205593109,
      "num_tokens": 937457.0,
      "step": 505
    },
    {
      "entropy": 7.911080026626587,
      "epoch": 0.04503708936771459,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002545,
      "loss": 7.6263,
      "mean_token_accuracy": 0.08180606141686439,
      "num_tokens": 946680.0,
      "step": 510
    },
    {
      "entropy": 7.8668982028961185,
      "epoch": 0.04547862945955493,
      "grad_norm": 0.90625,
      "learning_rate": 0.000257,
      "loss": 7.6101,
      "mean_token_accuracy": 0.08870264962315559,
      "num_tokens": 957088.0,
      "step": 515
    },
    {
      "entropy": 7.863427400588989,
      "epoch": 0.045920169551395267,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002595,
      "loss": 7.5106,
      "mean_token_accuracy": 0.08896929919719695,
      "num_tokens": 966171.0,
      "step": 520
    },
    {
      "entropy": 7.8639161586761475,
      "epoch": 0.046361709643235605,
      "grad_norm": 1.046875,
      "learning_rate": 0.000262,
      "loss": 7.4888,
      "mean_token_accuracy": 0.09330687001347542,
      "num_tokens": 975038.0,
      "step": 525
    },
    {
      "entropy": 7.818875646591186,
      "epoch": 0.04680324973507594,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.5302,
      "mean_token_accuracy": 0.08389112651348114,
      "num_tokens": 985559.0,
      "step": 530
    },
    {
      "entropy": 7.823013496398926,
      "epoch": 0.04724478982691628,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.5115,
      "mean_token_accuracy": 0.08914603590965271,
      "num_tokens": 994342.0,
      "step": 535
    },
    {
      "entropy": 7.819721317291259,
      "epoch": 0.047686329918756626,
      "grad_norm": 0.96875,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.4312,
      "mean_token_accuracy": 0.09510840475559235,
      "num_tokens": 1003246.0,
      "step": 540
    },
    {
      "entropy": 7.761527013778687,
      "epoch": 0.048127870010596964,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.4377,
      "mean_token_accuracy": 0.09536420628428459,
      "num_tokens": 1011949.0,
      "step": 545
    },
    {
      "entropy": 7.710439205169678,
      "epoch": 0.0485694101024373,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002745,
      "loss": 7.4041,
      "mean_token_accuracy": 0.09840207025408745,
      "num_tokens": 1020117.0,
      "step": 550
    },
    {
      "entropy": 7.790178346633911,
      "epoch": 0.04901095019427764,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000277,
      "loss": 7.5068,
      "mean_token_accuracy": 0.09245016798377037,
      "num_tokens": 1028917.0,
      "step": 555
    },
    {
      "entropy": 7.74592957496643,
      "epoch": 0.04945249028611798,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002795,
      "loss": 7.4332,
      "mean_token_accuracy": 0.098258488625288,
      "num_tokens": 1037542.0,
      "step": 560
    },
    {
      "entropy": 7.822019767761231,
      "epoch": 0.049894030377958316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.4483,
      "mean_token_accuracy": 0.09288977608084678,
      "num_tokens": 1046735.0,
      "step": 565
    },
    {
      "entropy": 7.696895027160645,
      "epoch": 0.050335570469798654,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002845,
      "loss": 7.4109,
      "mean_token_accuracy": 0.09220683723688125,
      "num_tokens": 1055306.0,
      "step": 570
    },
    {
      "entropy": 7.7618858337402346,
      "epoch": 0.050777110561639,
      "grad_norm": 1.078125,
      "learning_rate": 0.000287,
      "loss": 7.4532,
      "mean_token_accuracy": 0.10119536891579628,
      "num_tokens": 1064340.0,
      "step": 575
    },
    {
      "entropy": 7.708547353744507,
      "epoch": 0.05121865065347934,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002895,
      "loss": 7.4029,
      "mean_token_accuracy": 0.09979302212595939,
      "num_tokens": 1074235.0,
      "step": 580
    },
    {
      "entropy": 7.783248233795166,
      "epoch": 0.051660190745319676,
      "grad_norm": 0.90625,
      "learning_rate": 0.000292,
      "loss": 7.5004,
      "mean_token_accuracy": 0.09268677458167077,
      "num_tokens": 1083858.0,
      "step": 585
    },
    {
      "entropy": 7.698816013336182,
      "epoch": 0.052101730837160014,
      "grad_norm": 1.125,
      "learning_rate": 0.0002945,
      "loss": 7.4726,
      "mean_token_accuracy": 0.09486816674470902,
      "num_tokens": 1093674.0,
      "step": 590
    },
    {
      "entropy": 7.755134534835816,
      "epoch": 0.05254327092900035,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000297,
      "loss": 7.4248,
      "mean_token_accuracy": 0.09409366473555565,
      "num_tokens": 1103166.0,
      "step": 595
    },
    {
      "entropy": 7.736843490600586,
      "epoch": 0.05298481102084069,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002995,
      "loss": 7.3856,
      "mean_token_accuracy": 0.0958656646311283,
      "num_tokens": 1111458.0,
      "step": 600
    },
    {
      "entropy": 7.718952560424805,
      "epoch": 0.05342635111268103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000302,
      "loss": 7.408,
      "mean_token_accuracy": 0.09764600619673729,
      "num_tokens": 1120844.0,
      "step": 605
    },
    {
      "entropy": 7.711683893203736,
      "epoch": 0.05386789120452137,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003045,
      "loss": 7.468,
      "mean_token_accuracy": 0.09388890862464905,
      "num_tokens": 1129948.0,
      "step": 610
    },
    {
      "entropy": 7.6961164474487305,
      "epoch": 0.05430943129636171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000307,
      "loss": 7.4152,
      "mean_token_accuracy": 0.09394951239228248,
      "num_tokens": 1140142.0,
      "step": 615
    },
    {
      "entropy": 7.727436828613281,
      "epoch": 0.05475097138820205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003095,
      "loss": 7.4145,
      "mean_token_accuracy": 0.09343566671013832,
      "num_tokens": 1148428.0,
      "step": 620
    },
    {
      "entropy": 7.692458534240723,
      "epoch": 0.05519251148004239,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000312,
      "loss": 7.4136,
      "mean_token_accuracy": 0.10091176852583886,
      "num_tokens": 1157949.0,
      "step": 625
    },
    {
      "entropy": 7.679105138778686,
      "epoch": 0.055634051571882726,
      "grad_norm": 1.0,
      "learning_rate": 0.0003145,
      "loss": 7.3581,
      "mean_token_accuracy": 0.09939868301153183,
      "num_tokens": 1166492.0,
      "step": 630
    },
    {
      "entropy": 7.625578594207764,
      "epoch": 0.056075591663723064,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000317,
      "loss": 7.3769,
      "mean_token_accuracy": 0.09818158000707626,
      "num_tokens": 1176212.0,
      "step": 635
    },
    {
      "entropy": 7.703707695007324,
      "epoch": 0.0565171317555634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003195,
      "loss": 7.3927,
      "mean_token_accuracy": 0.09552990198135376,
      "num_tokens": 1185909.0,
      "step": 640
    },
    {
      "entropy": 7.680257797241211,
      "epoch": 0.05695867184740375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000322,
      "loss": 7.3432,
      "mean_token_accuracy": 0.10017694756388665,
      "num_tokens": 1196370.0,
      "step": 645
    },
    {
      "entropy": 7.708677196502686,
      "epoch": 0.057400211939244085,
      "grad_norm": 1.0,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.3425,
      "mean_token_accuracy": 0.09847476780414581,
      "num_tokens": 1205581.0,
      "step": 650
    },
    {
      "entropy": 7.593413877487182,
      "epoch": 0.05784175203108442,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.354,
      "mean_token_accuracy": 0.0941406324505806,
      "num_tokens": 1214437.0,
      "step": 655
    },
    {
      "entropy": 7.702506065368652,
      "epoch": 0.05828329212292476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.411,
      "mean_token_accuracy": 0.095302315056324,
      "num_tokens": 1224690.0,
      "step": 660
    },
    {
      "entropy": 7.615428924560547,
      "epoch": 0.0587248322147651,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.2941,
      "mean_token_accuracy": 0.09989068508148194,
      "num_tokens": 1234370.0,
      "step": 665
    },
    {
      "entropy": 7.657983112335205,
      "epoch": 0.05916637230660544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.3656,
      "mean_token_accuracy": 0.10043076574802398,
      "num_tokens": 1243801.0,
      "step": 670
    },
    {
      "entropy": 7.7026012420654295,
      "epoch": 0.05960791239844578,
      "grad_norm": 1.046875,
      "learning_rate": 0.000337,
      "loss": 7.4161,
      "mean_token_accuracy": 0.09777282178401947,
      "num_tokens": 1253750.0,
      "step": 675
    },
    {
      "entropy": 7.55789999961853,
      "epoch": 0.06004945249028612,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003395,
      "loss": 7.3337,
      "mean_token_accuracy": 0.09737052693963051,
      "num_tokens": 1262593.0,
      "step": 680
    },
    {
      "entropy": 7.5179671287536625,
      "epoch": 0.06049099258212646,
      "grad_norm": 1.046875,
      "learning_rate": 0.000342,
      "loss": 7.2582,
      "mean_token_accuracy": 0.10692062303423881,
      "num_tokens": 1271919.0,
      "step": 685
    },
    {
      "entropy": 7.656077432632446,
      "epoch": 0.0609325326739668,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.2333,
      "mean_token_accuracy": 0.09819802343845367,
      "num_tokens": 1281211.0,
      "step": 690
    },
    {
      "entropy": 7.640378093719482,
      "epoch": 0.061374072765807135,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000347,
      "loss": 7.266,
      "mean_token_accuracy": 0.1042046993970871,
      "num_tokens": 1290030.0,
      "step": 695
    },
    {
      "entropy": 7.5324629783630375,
      "epoch": 0.06181561285764747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003495,
      "loss": 7.3256,
      "mean_token_accuracy": 0.09880118295550347,
      "num_tokens": 1299182.0,
      "step": 700
    },
    {
      "entropy": 7.511147689819336,
      "epoch": 0.06225715294948781,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000352,
      "loss": 7.1888,
      "mean_token_accuracy": 0.11093291938304901,
      "num_tokens": 1307807.0,
      "step": 705
    },
    {
      "entropy": 7.603596162796021,
      "epoch": 0.06269869304132815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003545,
      "loss": 7.2998,
      "mean_token_accuracy": 0.09913515821099281,
      "num_tokens": 1317213.0,
      "step": 710
    },
    {
      "entropy": 7.583738040924072,
      "epoch": 0.0631402331331685,
      "grad_norm": 1.046875,
      "learning_rate": 0.000357,
      "loss": 7.2939,
      "mean_token_accuracy": 0.10248328670859337,
      "num_tokens": 1326712.0,
      "step": 715
    },
    {
      "entropy": 7.438777256011963,
      "epoch": 0.06358177322500883,
      "grad_norm": 1.0,
      "learning_rate": 0.0003595,
      "loss": 7.1652,
      "mean_token_accuracy": 0.10802278816699981,
      "num_tokens": 1335197.0,
      "step": 720
    },
    {
      "entropy": 7.493983316421509,
      "epoch": 0.06402331331684917,
      "grad_norm": 1.078125,
      "learning_rate": 0.000362,
      "loss": 7.2128,
      "mean_token_accuracy": 0.10637593865394593,
      "num_tokens": 1345184.0,
      "step": 725
    },
    {
      "entropy": 7.524926710128784,
      "epoch": 0.06446485340868952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003645,
      "loss": 7.1743,
      "mean_token_accuracy": 0.10707266703248024,
      "num_tokens": 1354035.0,
      "step": 730
    },
    {
      "entropy": 7.510273838043213,
      "epoch": 0.06490639350052985,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000367,
      "loss": 7.2745,
      "mean_token_accuracy": 0.10252953544259072,
      "num_tokens": 1363694.0,
      "step": 735
    },
    {
      "entropy": 7.540264368057251,
      "epoch": 0.06534793359237019,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003695,
      "loss": 7.2861,
      "mean_token_accuracy": 0.10127113163471221,
      "num_tokens": 1373213.0,
      "step": 740
    },
    {
      "entropy": 7.644721269607544,
      "epoch": 0.06578947368421052,
      "grad_norm": 1.1875,
      "learning_rate": 0.000372,
      "loss": 7.3536,
      "mean_token_accuracy": 0.09904789999127388,
      "num_tokens": 1382583.0,
      "step": 745
    },
    {
      "entropy": 7.567507886886597,
      "epoch": 0.06623101377605087,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003745,
      "loss": 7.2901,
      "mean_token_accuracy": 0.10179097279906273,
      "num_tokens": 1393179.0,
      "step": 750
    },
    {
      "entropy": 7.4339879035949705,
      "epoch": 0.0666725538678912,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000377,
      "loss": 7.2578,
      "mean_token_accuracy": 0.10336068943142891,
      "num_tokens": 1402710.0,
      "step": 755
    },
    {
      "entropy": 7.391937112808227,
      "epoch": 0.06711409395973154,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003795,
      "loss": 7.1319,
      "mean_token_accuracy": 0.10740135014057159,
      "num_tokens": 1411951.0,
      "step": 760
    },
    {
      "entropy": 7.566329479217529,
      "epoch": 0.06755563405157189,
      "grad_norm": 1.078125,
      "learning_rate": 0.000382,
      "loss": 7.228,
      "mean_token_accuracy": 0.10449767932295799,
      "num_tokens": 1420987.0,
      "step": 765
    },
    {
      "entropy": 7.435188674926758,
      "epoch": 0.06799717414341222,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003845,
      "loss": 7.2097,
      "mean_token_accuracy": 0.10870371311903,
      "num_tokens": 1429816.0,
      "step": 770
    },
    {
      "entropy": 7.429537582397461,
      "epoch": 0.06843871423525257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.2684,
      "mean_token_accuracy": 0.10217438042163848,
      "num_tokens": 1439734.0,
      "step": 775
    },
    {
      "entropy": 7.496665811538696,
      "epoch": 0.0688802543270929,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.1856,
      "mean_token_accuracy": 0.10607536807656288,
      "num_tokens": 1448666.0,
      "step": 780
    },
    {
      "entropy": 7.520268392562866,
      "epoch": 0.06932179441893324,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.2336,
      "mean_token_accuracy": 0.10618305131793022,
      "num_tokens": 1458754.0,
      "step": 785
    },
    {
      "entropy": 7.4070884704589846,
      "epoch": 0.06976333451077357,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.0997,
      "mean_token_accuracy": 0.11265210434794426,
      "num_tokens": 1468409.0,
      "step": 790
    },
    {
      "entropy": 7.468406248092651,
      "epoch": 0.07020487460261392,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.1527,
      "mean_token_accuracy": 0.10318700969219208,
      "num_tokens": 1477133.0,
      "step": 795
    },
    {
      "entropy": 7.417441082000733,
      "epoch": 0.07064641469445426,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003995,
      "loss": 7.2323,
      "mean_token_accuracy": 0.1052546851336956,
      "num_tokens": 1487215.0,
      "step": 800
    },
    {
      "entropy": 7.510851049423218,
      "epoch": 0.0710879547862946,
      "grad_norm": 1.21875,
      "learning_rate": 0.000402,
      "loss": 7.2308,
      "mean_token_accuracy": 0.10290714055299759,
      "num_tokens": 1496545.0,
      "step": 805
    },
    {
      "entropy": 7.4078692436218265,
      "epoch": 0.07152949487813494,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004045,
      "loss": 7.1474,
      "mean_token_accuracy": 0.11182119399309158,
      "num_tokens": 1505189.0,
      "step": 810
    },
    {
      "entropy": 7.373153114318848,
      "epoch": 0.07197103496997527,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.1244,
      "mean_token_accuracy": 0.11001503765583039,
      "num_tokens": 1513398.0,
      "step": 815
    },
    {
      "entropy": 7.427165269851685,
      "epoch": 0.07241257506181561,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004095,
      "loss": 7.1308,
      "mean_token_accuracy": 0.10737015083432197,
      "num_tokens": 1522712.0,
      "step": 820
    },
    {
      "entropy": 7.457416105270386,
      "epoch": 0.07285411515365595,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000412,
      "loss": 7.1819,
      "mean_token_accuracy": 0.10398382544517518,
      "num_tokens": 1532487.0,
      "step": 825
    },
    {
      "entropy": 7.382591438293457,
      "epoch": 0.07329565524549629,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004145,
      "loss": 7.1938,
      "mean_token_accuracy": 0.10898119583725929,
      "num_tokens": 1541888.0,
      "step": 830
    },
    {
      "entropy": 7.489791011810302,
      "epoch": 0.07373719533733664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000417,
      "loss": 7.2249,
      "mean_token_accuracy": 0.10024029538035392,
      "num_tokens": 1552170.0,
      "step": 835
    },
    {
      "entropy": 7.418872737884522,
      "epoch": 0.07417873542917697,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004195,
      "loss": 7.1548,
      "mean_token_accuracy": 0.10629647746682166,
      "num_tokens": 1562012.0,
      "step": 840
    },
    {
      "entropy": 7.322513008117676,
      "epoch": 0.07462027552101731,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000422,
      "loss": 7.0843,
      "mean_token_accuracy": 0.10927556678652764,
      "num_tokens": 1570150.0,
      "step": 845
    },
    {
      "entropy": 7.437068700790405,
      "epoch": 0.07506181561285764,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004245,
      "loss": 7.1004,
      "mean_token_accuracy": 0.11221965476870537,
      "num_tokens": 1579627.0,
      "step": 850
    },
    {
      "entropy": 7.385348701477051,
      "epoch": 0.07550335570469799,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000427,
      "loss": 7.2112,
      "mean_token_accuracy": 0.1068318523466587,
      "num_tokens": 1590059.0,
      "step": 855
    },
    {
      "entropy": 7.306034994125366,
      "epoch": 0.07594489579653832,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004295,
      "loss": 7.0794,
      "mean_token_accuracy": 0.10881896913051606,
      "num_tokens": 1599656.0,
      "step": 860
    },
    {
      "entropy": 7.4415796279907225,
      "epoch": 0.07638643588837866,
      "grad_norm": 1.03125,
      "learning_rate": 0.000432,
      "loss": 7.0527,
      "mean_token_accuracy": 0.10823039263486862,
      "num_tokens": 1608708.0,
      "step": 865
    },
    {
      "entropy": 7.342567300796508,
      "epoch": 0.07682797598021901,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004345,
      "loss": 7.1278,
      "mean_token_accuracy": 0.10357554033398628,
      "num_tokens": 1617467.0,
      "step": 870
    },
    {
      "entropy": 7.296669960021973,
      "epoch": 0.07726951607205934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000437,
      "loss": 7.0572,
      "mean_token_accuracy": 0.10800771564245223,
      "num_tokens": 1626656.0,
      "step": 875
    },
    {
      "entropy": 7.441175413131714,
      "epoch": 0.07771105616389969,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004395,
      "loss": 7.1039,
      "mean_token_accuracy": 0.10841156914830208,
      "num_tokens": 1635769.0,
      "step": 880
    },
    {
      "entropy": 7.383777809143067,
      "epoch": 0.07815259625574002,
      "grad_norm": 1.25,
      "learning_rate": 0.000442,
      "loss": 7.2128,
      "mean_token_accuracy": 0.10014851167798042,
      "num_tokens": 1645237.0,
      "step": 885
    },
    {
      "entropy": 7.236069011688232,
      "epoch": 0.07859413634758036,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004445,
      "loss": 7.0403,
      "mean_token_accuracy": 0.11353820711374282,
      "num_tokens": 1654266.0,
      "step": 890
    },
    {
      "entropy": 7.362179565429687,
      "epoch": 0.0790356764394207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000447,
      "loss": 7.1451,
      "mean_token_accuracy": 0.1029183603823185,
      "num_tokens": 1663409.0,
      "step": 895
    },
    {
      "entropy": 7.379905128479004,
      "epoch": 0.07947721653126104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.092,
      "mean_token_accuracy": 0.10835795775055886,
      "num_tokens": 1672664.0,
      "step": 900
    },
    {
      "entropy": 7.292973804473877,
      "epoch": 0.07991875662310138,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.2004,
      "mean_token_accuracy": 0.10515450984239579,
      "num_tokens": 1682656.0,
      "step": 905
    },
    {
      "entropy": 7.3837714195251465,
      "epoch": 0.08036029671494171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.0557,
      "mean_token_accuracy": 0.10702643021941186,
      "num_tokens": 1692155.0,
      "step": 910
    },
    {
      "entropy": 7.325821781158448,
      "epoch": 0.08080183680678206,
      "grad_norm": 1.0,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.0631,
      "mean_token_accuracy": 0.11022284105420113,
      "num_tokens": 1701347.0,
      "step": 915
    },
    {
      "entropy": 7.298824167251587,
      "epoch": 0.08124337689862239,
      "grad_norm": 1.25,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.0693,
      "mean_token_accuracy": 0.11432591453194618,
      "num_tokens": 1710110.0,
      "step": 920
    },
    {
      "entropy": 7.375689554214477,
      "epoch": 0.08168491699046274,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000462,
      "loss": 7.0914,
      "mean_token_accuracy": 0.10915563106536866,
      "num_tokens": 1718880.0,
      "step": 925
    },
    {
      "entropy": 7.270999765396118,
      "epoch": 0.08212645708230307,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004645,
      "loss": 7.0282,
      "mean_token_accuracy": 0.11133510172367096,
      "num_tokens": 1729164.0,
      "step": 930
    },
    {
      "entropy": 7.216585683822632,
      "epoch": 0.08256799717414341,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000467,
      "loss": 7.0806,
      "mean_token_accuracy": 0.11017340123653412,
      "num_tokens": 1738407.0,
      "step": 935
    },
    {
      "entropy": 7.264530849456787,
      "epoch": 0.08300953726598376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004695,
      "loss": 7.0866,
      "mean_token_accuracy": 0.11002494916319847,
      "num_tokens": 1747972.0,
      "step": 940
    },
    {
      "entropy": 7.352582788467407,
      "epoch": 0.08345107735782409,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000472,
      "loss": 7.0913,
      "mean_token_accuracy": 0.1055463895201683,
      "num_tokens": 1757597.0,
      "step": 945
    },
    {
      "entropy": 7.296703195571899,
      "epoch": 0.08389261744966443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004745,
      "loss": 7.0832,
      "mean_token_accuracy": 0.10835389271378518,
      "num_tokens": 1766905.0,
      "step": 950
    },
    {
      "entropy": 7.316994667053223,
      "epoch": 0.08433415754150476,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000477,
      "loss": 7.0214,
      "mean_token_accuracy": 0.11587357893586159,
      "num_tokens": 1776483.0,
      "step": 955
    },
    {
      "entropy": 7.267442893981934,
      "epoch": 0.08477569763334511,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004795,
      "loss": 7.1195,
      "mean_token_accuracy": 0.10487466976046562,
      "num_tokens": 1787050.0,
      "step": 960
    },
    {
      "entropy": 7.335643339157104,
      "epoch": 0.08521723772518544,
      "grad_norm": 1.1875,
      "learning_rate": 0.000482,
      "loss": 7.0617,
      "mean_token_accuracy": 0.11197662577033043,
      "num_tokens": 1796972.0,
      "step": 965
    },
    {
      "entropy": 7.191023874282837,
      "epoch": 0.08565877781702579,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845,
      "loss": 6.9551,
      "mean_token_accuracy": 0.1202144518494606,
      "num_tokens": 1804779.0,
      "step": 970
    },
    {
      "entropy": 7.229815483093262,
      "epoch": 0.08610031790886613,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000487,
      "loss": 7.001,
      "mean_token_accuracy": 0.1176213040947914,
      "num_tokens": 1813373.0,
      "step": 975
    },
    {
      "entropy": 7.1644312858581545,
      "epoch": 0.08654185800070646,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004895,
      "loss": 6.8935,
      "mean_token_accuracy": 0.12180616855621337,
      "num_tokens": 1822042.0,
      "step": 980
    },
    {
      "entropy": 7.214623975753784,
      "epoch": 0.08698339809254681,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492,
      "loss": 7.0036,
      "mean_token_accuracy": 0.11261652708053589,
      "num_tokens": 1831695.0,
      "step": 985
    },
    {
      "entropy": 7.246127367019653,
      "epoch": 0.08742493818438714,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004945,
      "loss": 7.0218,
      "mean_token_accuracy": 0.11053162068128586,
      "num_tokens": 1841035.0,
      "step": 990
    },
    {
      "entropy": 7.273807048797607,
      "epoch": 0.08786647827622748,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497,
      "loss": 7.0012,
      "mean_token_accuracy": 0.11223224252462387,
      "num_tokens": 1849808.0,
      "step": 995
    },
    {
      "entropy": 7.180046844482422,
      "epoch": 0.08830801836806781,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995,
      "loss": 7.0557,
      "mean_token_accuracy": 0.11089910492300988,
      "num_tokens": 1859583.0,
      "step": 1000
    },
    {
      "entropy": 7.183477878570557,
      "epoch": 0.08874955845990816,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499999998589561,
      "loss": 6.9748,
      "mean_token_accuracy": 0.11256436705589294,
      "num_tokens": 1869060.0,
      "step": 1005
    },
    {
      "entropy": 7.321033000946045,
      "epoch": 0.0891910985517485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999999928596523,
      "loss": 7.0264,
      "mean_token_accuracy": 0.11636429131031037,
      "num_tokens": 1878963.0,
      "step": 1010
    },
    {
      "entropy": 7.146453619003296,
      "epoch": 0.08963263864358884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999827221219,
      "loss": 6.9548,
      "mean_token_accuracy": 0.11412466391921043,
      "num_tokens": 1888681.0,
      "step": 1015
    },
    {
      "entropy": 7.116451168060303,
      "epoch": 0.09007417873542918,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999999681769696,
      "loss": 6.9771,
      "mean_token_accuracy": 0.11544483378529549,
      "num_tokens": 1897293.0,
      "step": 1020
    },
    {
      "entropy": 7.188655519485474,
      "epoch": 0.09051571882726951,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499999949224196,
      "loss": 7.0143,
      "mean_token_accuracy": 0.11129978001117706,
      "num_tokens": 1907243.0,
      "step": 1025
    },
    {
      "entropy": 7.212197875976562,
      "epoch": 0.09095725891910986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999258638013,
      "loss": 6.9686,
      "mean_token_accuracy": 0.1180865965783596,
      "num_tokens": 1916979.0,
      "step": 1030
    },
    {
      "entropy": 7.376511192321777,
      "epoch": 0.09139879901095019,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999998980957861,
      "loss": 7.0302,
      "mean_token_accuracy": 0.1100818432867527,
      "num_tokens": 1926151.0,
      "step": 1035
    },
    {
      "entropy": 7.1434496402740475,
      "epoch": 0.09184033910279053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999998659201508,
      "loss": 7.0049,
      "mean_token_accuracy": 0.11358301714062691,
      "num_tokens": 1935847.0,
      "step": 1040
    },
    {
      "entropy": 7.185492038726807,
      "epoch": 0.09228187919463088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999998293368961,
      "loss": 6.9572,
      "mean_token_accuracy": 0.12387590035796166,
      "num_tokens": 1945114.0,
      "step": 1045
    },
    {
      "entropy": 7.154706048965454,
      "epoch": 0.09272341928647121,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999997883460227,
      "loss": 6.9569,
      "mean_token_accuracy": 0.11779660508036613,
      "num_tokens": 1954924.0,
      "step": 1050
    },
    {
      "entropy": 7.140568017959595,
      "epoch": 0.09316495937831155,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999997429475314,
      "loss": 6.9131,
      "mean_token_accuracy": 0.11738472655415536,
      "num_tokens": 1963946.0,
      "step": 1055
    },
    {
      "entropy": 7.117366647720337,
      "epoch": 0.09360649947015189,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999996931414232,
      "loss": 6.927,
      "mean_token_accuracy": 0.12068335115909576,
      "num_tokens": 1973780.0,
      "step": 1060
    },
    {
      "entropy": 7.151812124252319,
      "epoch": 0.09404803956199223,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999996389276988,
      "loss": 6.9555,
      "mean_token_accuracy": 0.11808405518531799,
      "num_tokens": 1983561.0,
      "step": 1065
    },
    {
      "entropy": 7.215504932403564,
      "epoch": 0.09448957965383256,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999995803063596,
      "loss": 6.9481,
      "mean_token_accuracy": 0.11382859721779823,
      "num_tokens": 1993257.0,
      "step": 1070
    },
    {
      "entropy": 6.997649192810059,
      "epoch": 0.0949311197456729,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999995172774065,
      "loss": 6.8483,
      "mean_token_accuracy": 0.12370764091610909,
      "num_tokens": 2003391.0,
      "step": 1075
    },
    {
      "entropy": 7.239151048660278,
      "epoch": 0.09537265983751325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999994498408408,
      "loss": 6.9732,
      "mean_token_accuracy": 0.11102722734212875,
      "num_tokens": 2012004.0,
      "step": 1080
    },
    {
      "entropy": 7.1418920993804935,
      "epoch": 0.09581419992935358,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999993779966639,
      "loss": 6.9566,
      "mean_token_accuracy": 0.11490248888731003,
      "num_tokens": 2022033.0,
      "step": 1085
    },
    {
      "entropy": 7.171054220199585,
      "epoch": 0.09625574002119393,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999993017448771,
      "loss": 6.9893,
      "mean_token_accuracy": 0.11266493499279022,
      "num_tokens": 2031628.0,
      "step": 1090
    },
    {
      "entropy": 7.1874267578125,
      "epoch": 0.09669728011303426,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999992210854821,
      "loss": 6.9058,
      "mean_token_accuracy": 0.1091111071407795,
      "num_tokens": 2040339.0,
      "step": 1095
    },
    {
      "entropy": 7.08800573348999,
      "epoch": 0.0971388202048746,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999991360184801,
      "loss": 6.9265,
      "mean_token_accuracy": 0.11344177350401878,
      "num_tokens": 2050099.0,
      "step": 1100
    },
    {
      "entropy": 7.174203872680664,
      "epoch": 0.09758036029671494,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999990465438731,
      "loss": 6.9676,
      "mean_token_accuracy": 0.11609438061714172,
      "num_tokens": 2059572.0,
      "step": 1105
    },
    {
      "entropy": 7.087515354156494,
      "epoch": 0.09802190038855528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999989526616628,
      "loss": 6.9018,
      "mean_token_accuracy": 0.12604590877890587,
      "num_tokens": 2068449.0,
      "step": 1110
    },
    {
      "entropy": 7.112388658523559,
      "epoch": 0.09846344048039563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999988543718509,
      "loss": 6.8388,
      "mean_token_accuracy": 0.11249179318547249,
      "num_tokens": 2077508.0,
      "step": 1115
    },
    {
      "entropy": 7.071923446655274,
      "epoch": 0.09890498057223596,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999987516744394,
      "loss": 6.8896,
      "mean_token_accuracy": 0.12044401988387107,
      "num_tokens": 2087458.0,
      "step": 1120
    },
    {
      "entropy": 7.085772848129272,
      "epoch": 0.0993465206640763,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999986445694303,
      "loss": 7.0003,
      "mean_token_accuracy": 0.11157744228839875,
      "num_tokens": 2096966.0,
      "step": 1125
    },
    {
      "entropy": 7.1603165626525875,
      "epoch": 0.09978806075591663,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999985330568258,
      "loss": 6.984,
      "mean_token_accuracy": 0.11366051360964775,
      "num_tokens": 2106528.0,
      "step": 1130
    },
    {
      "entropy": 7.182817792892456,
      "epoch": 0.10022960084775698,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999984171366278,
      "loss": 6.9151,
      "mean_token_accuracy": 0.12007406130433082,
      "num_tokens": 2114582.0,
      "step": 1135
    },
    {
      "entropy": 7.036507177352905,
      "epoch": 0.10067114093959731,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499998296808839,
      "loss": 6.9229,
      "mean_token_accuracy": 0.11831944435834885,
      "num_tokens": 2125675.0,
      "step": 1140
    },
    {
      "entropy": 7.13743371963501,
      "epoch": 0.10111268103143765,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999981720734615,
      "loss": 6.9518,
      "mean_token_accuracy": 0.11637505069375038,
      "num_tokens": 2134696.0,
      "step": 1145
    },
    {
      "entropy": 7.1318199157714846,
      "epoch": 0.101554221123278,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999980429304977,
      "loss": 6.8227,
      "mean_token_accuracy": 0.11497059166431427,
      "num_tokens": 2144110.0,
      "step": 1150
    },
    {
      "entropy": 7.095803737640381,
      "epoch": 0.10199576121511833,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999979093799502,
      "loss": 6.8998,
      "mean_token_accuracy": 0.11766083538532257,
      "num_tokens": 2153059.0,
      "step": 1155
    },
    {
      "entropy": 7.164034080505371,
      "epoch": 0.10243730130695868,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999977714218217,
      "loss": 6.9923,
      "mean_token_accuracy": 0.11412862613797188,
      "num_tokens": 2162322.0,
      "step": 1160
    },
    {
      "entropy": 7.045686197280884,
      "epoch": 0.102878841398799,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999976290561147,
      "loss": 6.8552,
      "mean_token_accuracy": 0.11943022832274437,
      "num_tokens": 2170707.0,
      "step": 1165
    },
    {
      "entropy": 7.092855930328369,
      "epoch": 0.10332038149063935,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999974822828322,
      "loss": 6.8983,
      "mean_token_accuracy": 0.11820499077439309,
      "num_tokens": 2180134.0,
      "step": 1170
    },
    {
      "entropy": 7.0811927318573,
      "epoch": 0.10376192158247968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499997331101977,
      "loss": 6.8482,
      "mean_token_accuracy": 0.11699439510703087,
      "num_tokens": 2188277.0,
      "step": 1175
    },
    {
      "entropy": 7.0605254650115965,
      "epoch": 0.10420346167432003,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499997175513552,
      "loss": 6.8229,
      "mean_token_accuracy": 0.12035220712423325,
      "num_tokens": 2197380.0,
      "step": 1180
    },
    {
      "entropy": 7.096225452423096,
      "epoch": 0.10464500176616037,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999970155175603,
      "loss": 6.9042,
      "mean_token_accuracy": 0.1133291132748127,
      "num_tokens": 2207183.0,
      "step": 1185
    },
    {
      "entropy": 7.040783309936524,
      "epoch": 0.1050865418580007,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499996851114005,
      "loss": 6.8536,
      "mean_token_accuracy": 0.11778900921344757,
      "num_tokens": 2216686.0,
      "step": 1190
    },
    {
      "entropy": 7.115582895278931,
      "epoch": 0.10552808194984105,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999966823028894,
      "loss": 6.9219,
      "mean_token_accuracy": 0.11533070355653763,
      "num_tokens": 2226060.0,
      "step": 1195
    },
    {
      "entropy": 7.038763046264648,
      "epoch": 0.10596962204168138,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999965090842168,
      "loss": 6.8396,
      "mean_token_accuracy": 0.12296478524804115,
      "num_tokens": 2235941.0,
      "step": 1200
    },
    {
      "entropy": 6.94191164970398,
      "epoch": 0.10641116213352173,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999963314579905,
      "loss": 6.8194,
      "mean_token_accuracy": 0.12413551509380341,
      "num_tokens": 2245026.0,
      "step": 1205
    },
    {
      "entropy": 7.188455581665039,
      "epoch": 0.10685270222536206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999961494242139,
      "loss": 6.9298,
      "mean_token_accuracy": 0.11509394273161888,
      "num_tokens": 2254261.0,
      "step": 1210
    },
    {
      "entropy": 7.011233329772949,
      "epoch": 0.1072942423172024,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999959629828908,
      "loss": 6.8824,
      "mean_token_accuracy": 0.11670115143060684,
      "num_tokens": 2263025.0,
      "step": 1215
    },
    {
      "entropy": 7.086313724517822,
      "epoch": 0.10773578240904275,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999957721340248,
      "loss": 6.8529,
      "mean_token_accuracy": 0.1207992248237133,
      "num_tokens": 2271970.0,
      "step": 1220
    },
    {
      "entropy": 7.134052801132202,
      "epoch": 0.10817732250088308,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999955768776195,
      "loss": 6.8253,
      "mean_token_accuracy": 0.12035808637738228,
      "num_tokens": 2281624.0,
      "step": 1225
    },
    {
      "entropy": 6.951407861709595,
      "epoch": 0.10861886259272342,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999953772136788,
      "loss": 6.8166,
      "mean_token_accuracy": 0.1174696959555149,
      "num_tokens": 2290151.0,
      "step": 1230
    },
    {
      "entropy": 7.069948720932007,
      "epoch": 0.10906040268456375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999951731422068,
      "loss": 6.8155,
      "mean_token_accuracy": 0.12470634207129479,
      "num_tokens": 2299143.0,
      "step": 1235
    },
    {
      "entropy": 7.0021812438964846,
      "epoch": 0.1095019427764041,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999949646632072,
      "loss": 6.8169,
      "mean_token_accuracy": 0.11980563923716545,
      "num_tokens": 2308257.0,
      "step": 1240
    },
    {
      "entropy": 7.0178139209747314,
      "epoch": 0.10994348286824443,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999947517766842,
      "loss": 6.8092,
      "mean_token_accuracy": 0.12045684233307838,
      "num_tokens": 2318192.0,
      "step": 1245
    },
    {
      "entropy": 7.048533487319946,
      "epoch": 0.11038502296008477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499994534482642,
      "loss": 6.8642,
      "mean_token_accuracy": 0.11663642451167107,
      "num_tokens": 2327015.0,
      "step": 1250
    },
    {
      "entropy": 7.001744413375855,
      "epoch": 0.11082656305192512,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999943127810847,
      "loss": 6.8725,
      "mean_token_accuracy": 0.11562840417027473,
      "num_tokens": 2336873.0,
      "step": 1255
    },
    {
      "entropy": 7.011013555526733,
      "epoch": 0.11126810314376545,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999940866720169,
      "loss": 6.821,
      "mean_token_accuracy": 0.12255764603614808,
      "num_tokens": 2345193.0,
      "step": 1260
    },
    {
      "entropy": 7.060343503952026,
      "epoch": 0.1117096432356058,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999938561554429,
      "loss": 6.732,
      "mean_token_accuracy": 0.12168303355574608,
      "num_tokens": 2355368.0,
      "step": 1265
    },
    {
      "entropy": 6.952377223968506,
      "epoch": 0.11215118332744613,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999936212313672,
      "loss": 6.8565,
      "mean_token_accuracy": 0.11511669009923935,
      "num_tokens": 2365480.0,
      "step": 1270
    },
    {
      "entropy": 7.023001718521118,
      "epoch": 0.11259272341928647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999933818997943,
      "loss": 6.8131,
      "mean_token_accuracy": 0.1278669573366642,
      "num_tokens": 2373655.0,
      "step": 1275
    },
    {
      "entropy": 7.012344169616699,
      "epoch": 0.1130342635111268,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999931381607292,
      "loss": 6.8204,
      "mean_token_accuracy": 0.12402483075857162,
      "num_tokens": 2383283.0,
      "step": 1280
    },
    {
      "entropy": 7.009754276275634,
      "epoch": 0.11347580360296715,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999928900141764,
      "loss": 6.8781,
      "mean_token_accuracy": 0.11701322942972184,
      "num_tokens": 2392970.0,
      "step": 1285
    },
    {
      "entropy": 7.004397630691528,
      "epoch": 0.1139173436948075,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499992637460141,
      "loss": 6.7458,
      "mean_token_accuracy": 0.1333714246749878,
      "num_tokens": 2402594.0,
      "step": 1290
    },
    {
      "entropy": 6.943610906600952,
      "epoch": 0.11435888378664782,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999923804986275,
      "loss": 6.7477,
      "mean_token_accuracy": 0.12496615052223206,
      "num_tokens": 2411718.0,
      "step": 1295
    },
    {
      "entropy": 6.986547565460205,
      "epoch": 0.11480042387848817,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999921191296415,
      "loss": 6.8248,
      "mean_token_accuracy": 0.11757947653532028,
      "num_tokens": 2421285.0,
      "step": 1300
    },
    {
      "entropy": 7.024937295913697,
      "epoch": 0.1152419639703285,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999918533531877,
      "loss": 6.877,
      "mean_token_accuracy": 0.12012948319315911,
      "num_tokens": 2431070.0,
      "step": 1305
    },
    {
      "entropy": 6.981199169158936,
      "epoch": 0.11568350406216885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999915831692714,
      "loss": 6.7876,
      "mean_token_accuracy": 0.1231262855231762,
      "num_tokens": 2440572.0,
      "step": 1310
    },
    {
      "entropy": 6.893063688278199,
      "epoch": 0.11612504415400918,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999913085778981,
      "loss": 6.6922,
      "mean_token_accuracy": 0.12878676727414132,
      "num_tokens": 2449451.0,
      "step": 1315
    },
    {
      "entropy": 6.9775279521942135,
      "epoch": 0.11656658424584952,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999910295790729,
      "loss": 6.8187,
      "mean_token_accuracy": 0.11927035376429558,
      "num_tokens": 2458803.0,
      "step": 1320
    },
    {
      "entropy": 7.028124904632568,
      "epoch": 0.11700812433768987,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999907461728014,
      "loss": 6.7396,
      "mean_token_accuracy": 0.12259037122130394,
      "num_tokens": 2467838.0,
      "step": 1325
    },
    {
      "entropy": 6.956664562225342,
      "epoch": 0.1174496644295302,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999904583590893,
      "loss": 6.708,
      "mean_token_accuracy": 0.1303823284804821,
      "num_tokens": 2476726.0,
      "step": 1330
    },
    {
      "entropy": 6.849405431747437,
      "epoch": 0.11789120452137054,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999901661379418,
      "loss": 6.6171,
      "mean_token_accuracy": 0.13071923553943635,
      "num_tokens": 2485249.0,
      "step": 1335
    },
    {
      "entropy": 6.931700134277344,
      "epoch": 0.11833274461321087,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999898695093652,
      "loss": 6.7519,
      "mean_token_accuracy": 0.12411127239465714,
      "num_tokens": 2494136.0,
      "step": 1340
    },
    {
      "entropy": 6.999692344665528,
      "epoch": 0.11877428470505122,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999895684733648,
      "loss": 6.7524,
      "mean_token_accuracy": 0.12757925391197206,
      "num_tokens": 2504282.0,
      "step": 1345
    },
    {
      "entropy": 6.92269287109375,
      "epoch": 0.11921582479689156,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999892630299467,
      "loss": 6.7272,
      "mean_token_accuracy": 0.1256536178290844,
      "num_tokens": 2513313.0,
      "step": 1350
    },
    {
      "entropy": 6.8869856834411625,
      "epoch": 0.1196573648887319,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999889531791171,
      "loss": 6.7139,
      "mean_token_accuracy": 0.12457521334290504,
      "num_tokens": 2522824.0,
      "step": 1355
    },
    {
      "entropy": 6.899789142608642,
      "epoch": 0.12009890498057224,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999886389208817,
      "loss": 6.7113,
      "mean_token_accuracy": 0.12446566447615623,
      "num_tokens": 2531362.0,
      "step": 1360
    },
    {
      "entropy": 6.9725359916687015,
      "epoch": 0.12054044507241257,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999883202552468,
      "loss": 6.8036,
      "mean_token_accuracy": 0.11904620304703713,
      "num_tokens": 2540389.0,
      "step": 1365
    },
    {
      "entropy": 6.9280036926269535,
      "epoch": 0.12098198516425292,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999879971822189,
      "loss": 6.7275,
      "mean_token_accuracy": 0.12448851019144058,
      "num_tokens": 2550468.0,
      "step": 1370
    },
    {
      "entropy": 6.879989242553711,
      "epoch": 0.12142352525609325,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999876697018038,
      "loss": 6.7523,
      "mean_token_accuracy": 0.12376532703638077,
      "num_tokens": 2559998.0,
      "step": 1375
    },
    {
      "entropy": 6.906263637542724,
      "epoch": 0.1218650653479336,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999873378140085,
      "loss": 6.6884,
      "mean_token_accuracy": 0.12875040695071222,
      "num_tokens": 2570058.0,
      "step": 1380
    },
    {
      "entropy": 6.964244222640991,
      "epoch": 0.12230660543977394,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999870015188389,
      "loss": 6.693,
      "mean_token_accuracy": 0.12742569595575332,
      "num_tokens": 2578859.0,
      "step": 1385
    },
    {
      "entropy": 6.930225372314453,
      "epoch": 0.12274814553161427,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999866608163021,
      "loss": 6.7472,
      "mean_token_accuracy": 0.12182884812355041,
      "num_tokens": 2588639.0,
      "step": 1390
    },
    {
      "entropy": 6.946022987365723,
      "epoch": 0.12318968562345461,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999863157064045,
      "loss": 6.6954,
      "mean_token_accuracy": 0.13521388545632362,
      "num_tokens": 2597679.0,
      "step": 1395
    },
    {
      "entropy": 6.886091995239258,
      "epoch": 0.12363122571529495,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999859661891529,
      "loss": 6.7843,
      "mean_token_accuracy": 0.12181777730584145,
      "num_tokens": 2607841.0,
      "step": 1400
    },
    {
      "entropy": 6.958340835571289,
      "epoch": 0.12407276580713529,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999856122645543,
      "loss": 6.7371,
      "mean_token_accuracy": 0.12391316443681717,
      "num_tokens": 2618023.0,
      "step": 1405
    },
    {
      "entropy": 6.9667067527771,
      "epoch": 0.12451430589897562,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999852539326154,
      "loss": 6.7222,
      "mean_token_accuracy": 0.12619368433952333,
      "num_tokens": 2627174.0,
      "step": 1410
    },
    {
      "entropy": 6.849303531646728,
      "epoch": 0.12495584599081597,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999848911933434,
      "loss": 6.673,
      "mean_token_accuracy": 0.12527169808745384,
      "num_tokens": 2636485.0,
      "step": 1415
    },
    {
      "entropy": 6.840041828155518,
      "epoch": 0.1253973860826563,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999845240467453,
      "loss": 6.7288,
      "mean_token_accuracy": 0.12444792836904525,
      "num_tokens": 2645484.0,
      "step": 1420
    },
    {
      "entropy": 6.993687105178833,
      "epoch": 0.12583892617449666,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999841524928282,
      "loss": 6.6797,
      "mean_token_accuracy": 0.12360677868127823,
      "num_tokens": 2654753.0,
      "step": 1425
    },
    {
      "entropy": 6.829198551177979,
      "epoch": 0.126280466266337,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999837765315997,
      "loss": 6.749,
      "mean_token_accuracy": 0.12858281508088112,
      "num_tokens": 2663861.0,
      "step": 1430
    },
    {
      "entropy": 6.992751169204712,
      "epoch": 0.12672200635817732,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999833961630669,
      "loss": 6.7325,
      "mean_token_accuracy": 0.12685227245092393,
      "num_tokens": 2673426.0,
      "step": 1435
    },
    {
      "entropy": 6.908594751358033,
      "epoch": 0.12716354645001765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999830113872374,
      "loss": 6.7442,
      "mean_token_accuracy": 0.12368026599287987,
      "num_tokens": 2682597.0,
      "step": 1440
    },
    {
      "entropy": 6.829099655151367,
      "epoch": 0.127605086541858,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999826222041186,
      "loss": 6.6518,
      "mean_token_accuracy": 0.12994373813271523,
      "num_tokens": 2691299.0,
      "step": 1445
    },
    {
      "entropy": 6.969326400756836,
      "epoch": 0.12804662663369834,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999822286137182,
      "loss": 6.7546,
      "mean_token_accuracy": 0.12233989536762238,
      "num_tokens": 2701133.0,
      "step": 1450
    },
    {
      "entropy": 6.840322494506836,
      "epoch": 0.12848816672553867,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999818306160439,
      "loss": 6.7135,
      "mean_token_accuracy": 0.12351029440760612,
      "num_tokens": 2710704.0,
      "step": 1455
    },
    {
      "entropy": 6.927147150039673,
      "epoch": 0.12892970681737903,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999814282111034,
      "loss": 6.6843,
      "mean_token_accuracy": 0.12635136246681214,
      "num_tokens": 2719220.0,
      "step": 1460
    },
    {
      "entropy": 6.858917474746704,
      "epoch": 0.12937124690921936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999810213989047,
      "loss": 6.7173,
      "mean_token_accuracy": 0.1293303906917572,
      "num_tokens": 2728931.0,
      "step": 1465
    },
    {
      "entropy": 6.898620748519898,
      "epoch": 0.1298127870010597,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999806101794558,
      "loss": 6.739,
      "mean_token_accuracy": 0.12490534409880638,
      "num_tokens": 2738933.0,
      "step": 1470
    },
    {
      "entropy": 6.900324058532715,
      "epoch": 0.13025432709290002,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999801945527648,
      "loss": 6.6992,
      "mean_token_accuracy": 0.12481218501925469,
      "num_tokens": 2748807.0,
      "step": 1475
    },
    {
      "entropy": 6.84165096282959,
      "epoch": 0.13069586718474038,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999797745188395,
      "loss": 6.6427,
      "mean_token_accuracy": 0.12684471234679223,
      "num_tokens": 2757377.0,
      "step": 1480
    },
    {
      "entropy": 6.855244731903076,
      "epoch": 0.13113740727658071,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999793500776886,
      "loss": 6.7431,
      "mean_token_accuracy": 0.1237080566585064,
      "num_tokens": 2767621.0,
      "step": 1485
    },
    {
      "entropy": 6.912423419952392,
      "epoch": 0.13157894736842105,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999789212293201,
      "loss": 6.6604,
      "mean_token_accuracy": 0.12734709233045577,
      "num_tokens": 2776946.0,
      "step": 1490
    },
    {
      "entropy": 6.797172594070434,
      "epoch": 0.1320204874602614,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999784879737423,
      "loss": 6.6914,
      "mean_token_accuracy": 0.12315848618745803,
      "num_tokens": 2786027.0,
      "step": 1495
    },
    {
      "entropy": 6.886532497406006,
      "epoch": 0.13246202755210174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999780503109642,
      "loss": 6.6162,
      "mean_token_accuracy": 0.13158534690737725,
      "num_tokens": 2794131.0,
      "step": 1500
    },
    {
      "entropy": 6.920978593826294,
      "epoch": 0.13290356764394207,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999776082409939,
      "loss": 6.6941,
      "mean_token_accuracy": 0.12774772495031356,
      "num_tokens": 2803066.0,
      "step": 1505
    },
    {
      "entropy": 6.776230764389038,
      "epoch": 0.1333451077357824,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999771617638401,
      "loss": 6.5783,
      "mean_token_accuracy": 0.13168612942099572,
      "num_tokens": 2813321.0,
      "step": 1510
    },
    {
      "entropy": 6.877940082550049,
      "epoch": 0.13378664782762276,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999767108795118,
      "loss": 6.7411,
      "mean_token_accuracy": 0.12401887774467468,
      "num_tokens": 2822770.0,
      "step": 1515
    },
    {
      "entropy": 6.8076917171478275,
      "epoch": 0.1342281879194631,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999762555880176,
      "loss": 6.5542,
      "mean_token_accuracy": 0.1320797972381115,
      "num_tokens": 2830986.0,
      "step": 1520
    },
    {
      "entropy": 6.787933206558227,
      "epoch": 0.13466972801130342,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999757958893666,
      "loss": 6.7605,
      "mean_token_accuracy": 0.12231313586235046,
      "num_tokens": 2840808.0,
      "step": 1525
    },
    {
      "entropy": 6.998130750656128,
      "epoch": 0.13511126810314378,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004999753317835677,
      "loss": 6.6455,
      "mean_token_accuracy": 0.12353885173797607,
      "num_tokens": 2850043.0,
      "step": 1530
    },
    {
      "entropy": 6.652038717269898,
      "epoch": 0.1355528081949841,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999748632706299,
      "loss": 6.6144,
      "mean_token_accuracy": 0.12928197756409646,
      "num_tokens": 2860088.0,
      "step": 1535
    },
    {
      "entropy": 6.91050090789795,
      "epoch": 0.13599434828682444,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999743903505626,
      "loss": 6.6809,
      "mean_token_accuracy": 0.12998059689998626,
      "num_tokens": 2869564.0,
      "step": 1540
    },
    {
      "entropy": 6.814730072021485,
      "epoch": 0.13643588837866477,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999739130233749,
      "loss": 6.6792,
      "mean_token_accuracy": 0.12127165272831916,
      "num_tokens": 2878511.0,
      "step": 1545
    },
    {
      "entropy": 6.931462144851684,
      "epoch": 0.13687742847050513,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999734312890761,
      "loss": 6.6895,
      "mean_token_accuracy": 0.12624901607632638,
      "num_tokens": 2888424.0,
      "step": 1550
    },
    {
      "entropy": 6.782437467575074,
      "epoch": 0.13731896856234546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999729451476757,
      "loss": 6.6543,
      "mean_token_accuracy": 0.12920576557517052,
      "num_tokens": 2897636.0,
      "step": 1555
    },
    {
      "entropy": 6.844197607040405,
      "epoch": 0.1377605086541858,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999724545991835,
      "loss": 6.563,
      "mean_token_accuracy": 0.13266987800598146,
      "num_tokens": 2907132.0,
      "step": 1560
    },
    {
      "entropy": 6.810891151428223,
      "epoch": 0.13820204874602615,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999719596436086,
      "loss": 6.6234,
      "mean_token_accuracy": 0.12518846914172171,
      "num_tokens": 2916491.0,
      "step": 1565
    },
    {
      "entropy": 6.79849009513855,
      "epoch": 0.13864358883786648,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999714602809611,
      "loss": 6.5621,
      "mean_token_accuracy": 0.1363304764032364,
      "num_tokens": 2925656.0,
      "step": 1570
    },
    {
      "entropy": 6.793143224716187,
      "epoch": 0.1390851289297068,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999709565112506,
      "loss": 6.6186,
      "mean_token_accuracy": 0.13135875687003135,
      "num_tokens": 2934748.0,
      "step": 1575
    },
    {
      "entropy": 6.717667436599731,
      "epoch": 0.13952666902154715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499970448334487,
      "loss": 6.6221,
      "mean_token_accuracy": 0.12501492649316787,
      "num_tokens": 2943606.0,
      "step": 1580
    },
    {
      "entropy": 6.882297611236572,
      "epoch": 0.1399682091133875,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999699357506803,
      "loss": 6.6275,
      "mean_token_accuracy": 0.126065044850111,
      "num_tokens": 2952756.0,
      "step": 1585
    },
    {
      "entropy": 6.776244974136352,
      "epoch": 0.14040974920522784,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999694187598406,
      "loss": 6.5767,
      "mean_token_accuracy": 0.12857100814580918,
      "num_tokens": 2961790.0,
      "step": 1590
    },
    {
      "entropy": 6.833420181274414,
      "epoch": 0.14085128929706817,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999688973619777,
      "loss": 6.5744,
      "mean_token_accuracy": 0.12646075263619422,
      "num_tokens": 2970939.0,
      "step": 1595
    },
    {
      "entropy": 6.802111101150513,
      "epoch": 0.14129282938890853,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999683715571022,
      "loss": 6.6249,
      "mean_token_accuracy": 0.12900373488664627,
      "num_tokens": 2980182.0,
      "step": 1600
    },
    {
      "entropy": 6.9092512130737305,
      "epoch": 0.14173436948074886,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999678413452242,
      "loss": 6.6787,
      "mean_token_accuracy": 0.12571411207318306,
      "num_tokens": 2990119.0,
      "step": 1605
    },
    {
      "entropy": 6.721599006652832,
      "epoch": 0.1421759095725892,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999673067263542,
      "loss": 6.6058,
      "mean_token_accuracy": 0.13111452385783195,
      "num_tokens": 2999889.0,
      "step": 1610
    },
    {
      "entropy": 6.784647035598755,
      "epoch": 0.14261744966442952,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999667677005026,
      "loss": 6.5981,
      "mean_token_accuracy": 0.13359114453196524,
      "num_tokens": 3009204.0,
      "step": 1615
    },
    {
      "entropy": 6.784178304672241,
      "epoch": 0.14305898975626988,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999662242676799,
      "loss": 6.6844,
      "mean_token_accuracy": 0.1268974155187607,
      "num_tokens": 3018491.0,
      "step": 1620
    },
    {
      "entropy": 6.795537805557251,
      "epoch": 0.1435005298481102,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999656764278968,
      "loss": 6.6208,
      "mean_token_accuracy": 0.13830544129014016,
      "num_tokens": 3028330.0,
      "step": 1625
    },
    {
      "entropy": 6.7406950950622555,
      "epoch": 0.14394206993995054,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999651241811642,
      "loss": 6.5979,
      "mean_token_accuracy": 0.12952237874269484,
      "num_tokens": 3038035.0,
      "step": 1630
    },
    {
      "entropy": 6.8958698272705075,
      "epoch": 0.1443836100317909,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999645675274925,
      "loss": 6.6833,
      "mean_token_accuracy": 0.12506106793880462,
      "num_tokens": 3047270.0,
      "step": 1635
    },
    {
      "entropy": 6.712257814407349,
      "epoch": 0.14482515012363123,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999640064668931,
      "loss": 6.5277,
      "mean_token_accuracy": 0.1350895307958126,
      "num_tokens": 3056540.0,
      "step": 1640
    },
    {
      "entropy": 6.760552787780762,
      "epoch": 0.14526669021547156,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999634409993766,
      "loss": 6.631,
      "mean_token_accuracy": 0.12776513770222664,
      "num_tokens": 3065800.0,
      "step": 1645
    },
    {
      "entropy": 6.860488700866699,
      "epoch": 0.1457082303073119,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999628711249544,
      "loss": 6.5654,
      "mean_token_accuracy": 0.13428230658173562,
      "num_tokens": 3075339.0,
      "step": 1650
    },
    {
      "entropy": 6.753561735153198,
      "epoch": 0.14614977039915225,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999622968436373,
      "loss": 6.6494,
      "mean_token_accuracy": 0.12430301308631897,
      "num_tokens": 3084798.0,
      "step": 1655
    },
    {
      "entropy": 6.865996789932251,
      "epoch": 0.14659131049099258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999617181554369,
      "loss": 6.6611,
      "mean_token_accuracy": 0.13118176385760308,
      "num_tokens": 3093437.0,
      "step": 1660
    },
    {
      "entropy": 6.732730865478516,
      "epoch": 0.1470328505828329,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999611350603643,
      "loss": 6.6614,
      "mean_token_accuracy": 0.125968486815691,
      "num_tokens": 3102588.0,
      "step": 1665
    },
    {
      "entropy": 6.854873085021973,
      "epoch": 0.14747439067467327,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000499960547558431,
      "loss": 6.546,
      "mean_token_accuracy": 0.13650740087032318,
      "num_tokens": 3112448.0,
      "step": 1670
    },
    {
      "entropy": 6.792056465148926,
      "epoch": 0.1479159307665136,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999599556496486,
      "loss": 6.6032,
      "mean_token_accuracy": 0.13347120061516762,
      "num_tokens": 3121760.0,
      "step": 1675
    },
    {
      "entropy": 6.710882043838501,
      "epoch": 0.14835747085835393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999593593340286,
      "loss": 6.5476,
      "mean_token_accuracy": 0.13694098070263863,
      "num_tokens": 3130585.0,
      "step": 1680
    },
    {
      "entropy": 6.756034183502197,
      "epoch": 0.14879901095019427,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999587586115826,
      "loss": 6.5101,
      "mean_token_accuracy": 0.1322891369462013,
      "num_tokens": 3139811.0,
      "step": 1685
    },
    {
      "entropy": 6.741791248321533,
      "epoch": 0.14924055104203462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999581534823226,
      "loss": 6.5782,
      "mean_token_accuracy": 0.13344234824180604,
      "num_tokens": 3148601.0,
      "step": 1690
    },
    {
      "entropy": 6.74654541015625,
      "epoch": 0.14968209113387496,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999575439462601,
      "loss": 6.5475,
      "mean_token_accuracy": 0.13243092596530914,
      "num_tokens": 3157639.0,
      "step": 1695
    },
    {
      "entropy": 6.863055038452148,
      "epoch": 0.1501236312257153,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999569300034075,
      "loss": 6.6686,
      "mean_token_accuracy": 0.12570775672793388,
      "num_tokens": 3168161.0,
      "step": 1700
    },
    {
      "entropy": 6.697668886184692,
      "epoch": 0.15056517131755565,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999563116537764,
      "loss": 6.5863,
      "mean_token_accuracy": 0.13283494636416435,
      "num_tokens": 3177388.0,
      "step": 1705
    },
    {
      "entropy": 6.749922609329223,
      "epoch": 0.15100671140939598,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999556888973792,
      "loss": 6.5652,
      "mean_token_accuracy": 0.13535942062735556,
      "num_tokens": 3187166.0,
      "step": 1710
    },
    {
      "entropy": 6.7436247825622555,
      "epoch": 0.1514482515012363,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999550617342279,
      "loss": 6.5117,
      "mean_token_accuracy": 0.13675145283341408,
      "num_tokens": 3197325.0,
      "step": 1715
    },
    {
      "entropy": 6.722959423065186,
      "epoch": 0.15188979159307664,
      "grad_norm": 1.0,
      "learning_rate": 0.000499954430164335,
      "loss": 6.5838,
      "mean_token_accuracy": 0.13445114344358444,
      "num_tokens": 3207348.0,
      "step": 1720
    },
    {
      "entropy": 6.767676830291748,
      "epoch": 0.152331331684917,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999537941877127,
      "loss": 6.5839,
      "mean_token_accuracy": 0.13400614261627197,
      "num_tokens": 3217291.0,
      "step": 1725
    },
    {
      "entropy": 6.651678419113159,
      "epoch": 0.15277287177675733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999531538043735,
      "loss": 6.4987,
      "mean_token_accuracy": 0.1366315647959709,
      "num_tokens": 3225427.0,
      "step": 1730
    },
    {
      "entropy": 6.813160705566406,
      "epoch": 0.15321441186859766,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999525090143298,
      "loss": 6.6344,
      "mean_token_accuracy": 0.13224097862839698,
      "num_tokens": 3235037.0,
      "step": 1735
    },
    {
      "entropy": 6.71365065574646,
      "epoch": 0.15365595196043802,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999518598175946,
      "loss": 6.4877,
      "mean_token_accuracy": 0.14198092222213746,
      "num_tokens": 3244318.0,
      "step": 1740
    },
    {
      "entropy": 6.645174169540406,
      "epoch": 0.15409749205227835,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999512062141805,
      "loss": 6.5706,
      "mean_token_accuracy": 0.13333735913038253,
      "num_tokens": 3253753.0,
      "step": 1745
    },
    {
      "entropy": 6.7497145652771,
      "epoch": 0.15453903214411868,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999505482040999,
      "loss": 6.4393,
      "mean_token_accuracy": 0.14499711841344834,
      "num_tokens": 3263081.0,
      "step": 1750
    },
    {
      "entropy": 6.651862573623657,
      "epoch": 0.154980572235959,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999498857873662,
      "loss": 6.5053,
      "mean_token_accuracy": 0.13713117241859435,
      "num_tokens": 3271091.0,
      "step": 1755
    },
    {
      "entropy": 6.762651491165161,
      "epoch": 0.15542211232779937,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999492189639921,
      "loss": 6.5627,
      "mean_token_accuracy": 0.13147899731993676,
      "num_tokens": 3280149.0,
      "step": 1760
    },
    {
      "entropy": 6.781657361984253,
      "epoch": 0.1558636524196397,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999485477339907,
      "loss": 6.4859,
      "mean_token_accuracy": 0.13011387065052987,
      "num_tokens": 3288644.0,
      "step": 1765
    },
    {
      "entropy": 6.6499279022216795,
      "epoch": 0.15630519251148003,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999478720973753,
      "loss": 6.5944,
      "mean_token_accuracy": 0.12865517288446426,
      "num_tokens": 3298756.0,
      "step": 1770
    },
    {
      "entropy": 6.82049560546875,
      "epoch": 0.1567467326033204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499947192054159,
      "loss": 6.572,
      "mean_token_accuracy": 0.12759956419467927,
      "num_tokens": 3307004.0,
      "step": 1775
    },
    {
      "entropy": 6.729887390136719,
      "epoch": 0.15718827269516072,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499946507604355,
      "loss": 6.525,
      "mean_token_accuracy": 0.13568256571888923,
      "num_tokens": 3316341.0,
      "step": 1780
    },
    {
      "entropy": 6.621280479431152,
      "epoch": 0.15762981278700106,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499945818747977,
      "loss": 6.4272,
      "mean_token_accuracy": 0.14301357716321944,
      "num_tokens": 3324837.0,
      "step": 1785
    },
    {
      "entropy": 6.751738595962524,
      "epoch": 0.1580713528788414,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999451254850383,
      "loss": 6.632,
      "mean_token_accuracy": 0.1333488866686821,
      "num_tokens": 3334029.0,
      "step": 1790
    },
    {
      "entropy": 6.715813636779785,
      "epoch": 0.15851289297068175,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999444278155525,
      "loss": 6.4757,
      "mean_token_accuracy": 0.13364549726247787,
      "num_tokens": 3343566.0,
      "step": 1795
    },
    {
      "entropy": 6.723165607452392,
      "epoch": 0.15895443306252208,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999437257395333,
      "loss": 6.5602,
      "mean_token_accuracy": 0.12862215787172318,
      "num_tokens": 3352428.0,
      "step": 1800
    },
    {
      "entropy": 6.686942768096924,
      "epoch": 0.1593959731543624,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999430192569944,
      "loss": 6.4482,
      "mean_token_accuracy": 0.13607168048620225,
      "num_tokens": 3361404.0,
      "step": 1805
    },
    {
      "entropy": 6.660811424255371,
      "epoch": 0.15983751324620277,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999423083679498,
      "loss": 6.5889,
      "mean_token_accuracy": 0.13365302309393884,
      "num_tokens": 3370341.0,
      "step": 1810
    },
    {
      "entropy": 6.763469409942627,
      "epoch": 0.1602790533380431,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999415930724133,
      "loss": 6.5229,
      "mean_token_accuracy": 0.1372690588235855,
      "num_tokens": 3379441.0,
      "step": 1815
    },
    {
      "entropy": 6.6314574718475345,
      "epoch": 0.16072059342988343,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999408733703988,
      "loss": 6.5031,
      "mean_token_accuracy": 0.1369751662015915,
      "num_tokens": 3389000.0,
      "step": 1820
    },
    {
      "entropy": 6.7453086376190186,
      "epoch": 0.16116213352172376,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999401492619207,
      "loss": 6.558,
      "mean_token_accuracy": 0.13744618371129036,
      "num_tokens": 3398897.0,
      "step": 1825
    },
    {
      "entropy": 6.6991571426391605,
      "epoch": 0.16160367361356412,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999394207469928,
      "loss": 6.5246,
      "mean_token_accuracy": 0.13513347804546355,
      "num_tokens": 3408234.0,
      "step": 1830
    },
    {
      "entropy": 6.602311992645264,
      "epoch": 0.16204521370540445,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999386878256297,
      "loss": 6.5632,
      "mean_token_accuracy": 0.1290775217115879,
      "num_tokens": 3417966.0,
      "step": 1835
    },
    {
      "entropy": 6.755477428436279,
      "epoch": 0.16248675379724478,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999379504978457,
      "loss": 6.5486,
      "mean_token_accuracy": 0.13179300874471664,
      "num_tokens": 3426689.0,
      "step": 1840
    },
    {
      "entropy": 6.736843299865723,
      "epoch": 0.16292829388908514,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499937208763655,
      "loss": 6.5904,
      "mean_token_accuracy": 0.13453762903809546,
      "num_tokens": 3437066.0,
      "step": 1845
    },
    {
      "entropy": 6.701791572570801,
      "epoch": 0.16336983398092547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999364626230724,
      "loss": 6.5238,
      "mean_token_accuracy": 0.1287774458527565,
      "num_tokens": 3446215.0,
      "step": 1850
    },
    {
      "entropy": 6.76169638633728,
      "epoch": 0.1638113740727658,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999357120761124,
      "loss": 6.4623,
      "mean_token_accuracy": 0.1368204079568386,
      "num_tokens": 3456173.0,
      "step": 1855
    },
    {
      "entropy": 6.623220682144165,
      "epoch": 0.16425291416460613,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999349571227898,
      "loss": 6.5465,
      "mean_token_accuracy": 0.1315374918282032,
      "num_tokens": 3465815.0,
      "step": 1860
    },
    {
      "entropy": 6.729666709899902,
      "epoch": 0.1646944542564465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999341977631193,
      "loss": 6.473,
      "mean_token_accuracy": 0.13990226536989211,
      "num_tokens": 3474396.0,
      "step": 1865
    },
    {
      "entropy": 6.622664070129394,
      "epoch": 0.16513599434828682,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999334339971157,
      "loss": 6.5267,
      "mean_token_accuracy": 0.1300884112715721,
      "num_tokens": 3484240.0,
      "step": 1870
    },
    {
      "entropy": 6.753061437606812,
      "epoch": 0.16557753444012716,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999326658247942,
      "loss": 6.4985,
      "mean_token_accuracy": 0.1349583201110363,
      "num_tokens": 3493569.0,
      "step": 1875
    },
    {
      "entropy": 6.659213352203369,
      "epoch": 0.16601907453196751,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999318932461696,
      "loss": 6.4129,
      "mean_token_accuracy": 0.12922252491116523,
      "num_tokens": 3501947.0,
      "step": 1880
    },
    {
      "entropy": 6.578718376159668,
      "epoch": 0.16646061462380785,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999311162612571,
      "loss": 6.4492,
      "mean_token_accuracy": 0.14246295243501664,
      "num_tokens": 3511817.0,
      "step": 1885
    },
    {
      "entropy": 6.711222791671753,
      "epoch": 0.16690215471564818,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499930334870072,
      "loss": 6.5397,
      "mean_token_accuracy": 0.1377376690506935,
      "num_tokens": 3521581.0,
      "step": 1890
    },
    {
      "entropy": 6.6527848720550535,
      "epoch": 0.1673436948074885,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999295490726296,
      "loss": 6.5204,
      "mean_token_accuracy": 0.13453154340386392,
      "num_tokens": 3531659.0,
      "step": 1895
    },
    {
      "entropy": 6.594851064682007,
      "epoch": 0.16778523489932887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999287588689453,
      "loss": 6.3853,
      "mean_token_accuracy": 0.14152303338050842,
      "num_tokens": 3540919.0,
      "step": 1900
    },
    {
      "entropy": 6.660305786132812,
      "epoch": 0.1682267749911692,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999279642590344,
      "loss": 6.3887,
      "mean_token_accuracy": 0.13910981193184851,
      "num_tokens": 3549756.0,
      "step": 1905
    },
    {
      "entropy": 6.607125902175904,
      "epoch": 0.16866831508300953,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999271652429127,
      "loss": 6.454,
      "mean_token_accuracy": 0.13921822756528854,
      "num_tokens": 3559686.0,
      "step": 1910
    },
    {
      "entropy": 6.609749174118042,
      "epoch": 0.1691098551748499,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999263618205958,
      "loss": 6.4051,
      "mean_token_accuracy": 0.13996847197413445,
      "num_tokens": 3568558.0,
      "step": 1915
    },
    {
      "entropy": 6.624215602874756,
      "epoch": 0.16955139526669022,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999255539920993,
      "loss": 6.484,
      "mean_token_accuracy": 0.1342615231871605,
      "num_tokens": 3578297.0,
      "step": 1920
    },
    {
      "entropy": 6.698274326324463,
      "epoch": 0.16999293535853055,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999247417574391,
      "loss": 6.5244,
      "mean_token_accuracy": 0.13907946199178695,
      "num_tokens": 3588372.0,
      "step": 1925
    },
    {
      "entropy": 6.692286682128906,
      "epoch": 0.17043447545037088,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999239251166312,
      "loss": 6.4649,
      "mean_token_accuracy": 0.14157049655914306,
      "num_tokens": 3597728.0,
      "step": 1930
    },
    {
      "entropy": 6.6316468715667725,
      "epoch": 0.17087601554221124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999231040696914,
      "loss": 6.454,
      "mean_token_accuracy": 0.14170244634151458,
      "num_tokens": 3606886.0,
      "step": 1935
    },
    {
      "entropy": 6.678344964981079,
      "epoch": 0.17131755563405157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999222786166361,
      "loss": 6.4845,
      "mean_token_accuracy": 0.13350907787680627,
      "num_tokens": 3615795.0,
      "step": 1940
    },
    {
      "entropy": 6.665940523147583,
      "epoch": 0.1717590957258919,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999214487574812,
      "loss": 6.4825,
      "mean_token_accuracy": 0.13486144170165063,
      "num_tokens": 3624360.0,
      "step": 1945
    },
    {
      "entropy": 6.654449462890625,
      "epoch": 0.17220063581773226,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999206144922431,
      "loss": 6.4428,
      "mean_token_accuracy": 0.1393147215247154,
      "num_tokens": 3633822.0,
      "step": 1950
    },
    {
      "entropy": 6.6238181591033936,
      "epoch": 0.1726421759095726,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499919775820938,
      "loss": 6.5075,
      "mean_token_accuracy": 0.13793846368789672,
      "num_tokens": 3643667.0,
      "step": 1955
    },
    {
      "entropy": 6.619096565246582,
      "epoch": 0.17308371600141292,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999189327435825,
      "loss": 6.4087,
      "mean_token_accuracy": 0.13899273052811623,
      "num_tokens": 3652596.0,
      "step": 1960
    },
    {
      "entropy": 6.695830154418945,
      "epoch": 0.17352525609325326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999180852601929,
      "loss": 6.5268,
      "mean_token_accuracy": 0.13822980970144272,
      "num_tokens": 3661362.0,
      "step": 1965
    },
    {
      "entropy": 6.632474565505982,
      "epoch": 0.17396679618509361,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499917233370786,
      "loss": 6.4865,
      "mean_token_accuracy": 0.1420795053243637,
      "num_tokens": 3671573.0,
      "step": 1970
    },
    {
      "entropy": 6.656900548934937,
      "epoch": 0.17440833627693395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999163770753784,
      "loss": 6.5884,
      "mean_token_accuracy": 0.1283446244895458,
      "num_tokens": 3681183.0,
      "step": 1975
    },
    {
      "entropy": 6.667371845245361,
      "epoch": 0.17484987636877428,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999155163739869,
      "loss": 6.4293,
      "mean_token_accuracy": 0.13852844312787055,
      "num_tokens": 3689768.0,
      "step": 1980
    },
    {
      "entropy": 6.64871244430542,
      "epoch": 0.17529141646061464,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999146512666284,
      "loss": 6.4866,
      "mean_token_accuracy": 0.13745958730578423,
      "num_tokens": 3699291.0,
      "step": 1985
    },
    {
      "entropy": 6.74596848487854,
      "epoch": 0.17573295655245497,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999137817533197,
      "loss": 6.4329,
      "mean_token_accuracy": 0.13396958485245705,
      "num_tokens": 3708551.0,
      "step": 1990
    },
    {
      "entropy": 6.543181896209717,
      "epoch": 0.1761744966442953,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999129078340779,
      "loss": 6.4554,
      "mean_token_accuracy": 0.13674973249435424,
      "num_tokens": 3717617.0,
      "step": 1995
    },
    {
      "entropy": 6.594240331649781,
      "epoch": 0.17661603673613563,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999120295089202,
      "loss": 6.4395,
      "mean_token_accuracy": 0.14458404257893562,
      "num_tokens": 3728457.0,
      "step": 2000
    },
    {
      "entropy": 6.636260986328125,
      "epoch": 0.177057576827976,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999111467778639,
      "loss": 6.5277,
      "mean_token_accuracy": 0.13387356027960778,
      "num_tokens": 3738467.0,
      "step": 2005
    },
    {
      "entropy": 6.71912145614624,
      "epoch": 0.17749911691981632,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499910259640926,
      "loss": 6.532,
      "mean_token_accuracy": 0.13334558084607123,
      "num_tokens": 3747482.0,
      "step": 2010
    },
    {
      "entropy": 6.6471014499664305,
      "epoch": 0.17794065701165665,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499909368098124,
      "loss": 6.5043,
      "mean_token_accuracy": 0.1343577802181244,
      "num_tokens": 3756227.0,
      "step": 2015
    },
    {
      "entropy": 6.543698358535766,
      "epoch": 0.178382197103497,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999084721494754,
      "loss": 6.4101,
      "mean_token_accuracy": 0.13692991584539413,
      "num_tokens": 3765481.0,
      "step": 2020
    },
    {
      "entropy": 6.616260051727295,
      "epoch": 0.17882373719533734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999075717949978,
      "loss": 6.4347,
      "mean_token_accuracy": 0.13827534690499305,
      "num_tokens": 3774261.0,
      "step": 2025
    },
    {
      "entropy": 6.604419136047364,
      "epoch": 0.17926527728717767,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999066670347089,
      "loss": 6.4235,
      "mean_token_accuracy": 0.14103795886039733,
      "num_tokens": 3783725.0,
      "step": 2030
    },
    {
      "entropy": 6.695284843444824,
      "epoch": 0.179706817379018,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004999057578686261,
      "loss": 6.5538,
      "mean_token_accuracy": 0.13175728917121887,
      "num_tokens": 3794129.0,
      "step": 2035
    },
    {
      "entropy": 6.723241186141967,
      "epoch": 0.18014835747085836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999048442967675,
      "loss": 6.5037,
      "mean_token_accuracy": 0.13313911259174346,
      "num_tokens": 3803954.0,
      "step": 2040
    },
    {
      "entropy": 6.576015520095825,
      "epoch": 0.1805898975626987,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999039263191508,
      "loss": 6.4143,
      "mean_token_accuracy": 0.13729096055030823,
      "num_tokens": 3812220.0,
      "step": 2045
    },
    {
      "entropy": 6.613183689117432,
      "epoch": 0.18103143765453902,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999030039357943,
      "loss": 6.4089,
      "mean_token_accuracy": 0.1392333209514618,
      "num_tokens": 3821595.0,
      "step": 2050
    },
    {
      "entropy": 6.618626260757447,
      "epoch": 0.18147297774637938,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999020771467158,
      "loss": 6.4898,
      "mean_token_accuracy": 0.13633593022823334,
      "num_tokens": 3831020.0,
      "step": 2055
    },
    {
      "entropy": 6.593099689483642,
      "epoch": 0.1819145178382197,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999011459519335,
      "loss": 6.4288,
      "mean_token_accuracy": 0.13460941165685653,
      "num_tokens": 3839735.0,
      "step": 2060
    },
    {
      "entropy": 6.643797731399536,
      "epoch": 0.18235605793006004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999002103514655,
      "loss": 6.5236,
      "mean_token_accuracy": 0.13866995573043822,
      "num_tokens": 3849095.0,
      "step": 2065
    },
    {
      "entropy": 6.676905632019043,
      "epoch": 0.18279759802190038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998992703453304,
      "loss": 6.5015,
      "mean_token_accuracy": 0.14171259924769403,
      "num_tokens": 3859063.0,
      "step": 2070
    },
    {
      "entropy": 6.639267778396606,
      "epoch": 0.18323913811374073,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998983259335466,
      "loss": 6.5092,
      "mean_token_accuracy": 0.13027978986501693,
      "num_tokens": 3868626.0,
      "step": 2075
    },
    {
      "entropy": 6.607439041137695,
      "epoch": 0.18368067820558107,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998973771161324,
      "loss": 6.4954,
      "mean_token_accuracy": 0.139207723736763,
      "num_tokens": 3878477.0,
      "step": 2080
    },
    {
      "entropy": 6.611793661117554,
      "epoch": 0.1841222182974214,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998964238931065,
      "loss": 6.3776,
      "mean_token_accuracy": 0.14397683441638948,
      "num_tokens": 3887166.0,
      "step": 2085
    },
    {
      "entropy": 6.496103525161743,
      "epoch": 0.18456375838926176,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998954662644876,
      "loss": 6.3283,
      "mean_token_accuracy": 0.1425325997173786,
      "num_tokens": 3896431.0,
      "step": 2090
    },
    {
      "entropy": 6.614110946655273,
      "epoch": 0.1850052984811021,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998945042302943,
      "loss": 6.402,
      "mean_token_accuracy": 0.1468320146203041,
      "num_tokens": 3904937.0,
      "step": 2095
    },
    {
      "entropy": 6.649114370346069,
      "epoch": 0.18544683857294242,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998935377905457,
      "loss": 6.4103,
      "mean_token_accuracy": 0.14189980030059815,
      "num_tokens": 3913905.0,
      "step": 2100
    },
    {
      "entropy": 6.551333093643189,
      "epoch": 0.18588837866478275,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998925669452605,
      "loss": 6.4386,
      "mean_token_accuracy": 0.1394965521991253,
      "num_tokens": 3923894.0,
      "step": 2105
    },
    {
      "entropy": 6.607130718231201,
      "epoch": 0.1863299187566231,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998915916944579,
      "loss": 6.4454,
      "mean_token_accuracy": 0.1338847316801548,
      "num_tokens": 3932124.0,
      "step": 2110
    },
    {
      "entropy": 6.626301956176758,
      "epoch": 0.18677145884846344,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998906120381568,
      "loss": 6.4752,
      "mean_token_accuracy": 0.1314583897590637,
      "num_tokens": 3942462.0,
      "step": 2115
    },
    {
      "entropy": 6.55535740852356,
      "epoch": 0.18721299894030377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998896279763766,
      "loss": 6.4285,
      "mean_token_accuracy": 0.143652606010437,
      "num_tokens": 3951361.0,
      "step": 2120
    },
    {
      "entropy": 6.699276733398437,
      "epoch": 0.18765453903214413,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998886395091365,
      "loss": 6.5322,
      "mean_token_accuracy": 0.13593855649232864,
      "num_tokens": 3961486.0,
      "step": 2125
    },
    {
      "entropy": 6.510635089874268,
      "epoch": 0.18809607912398446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998876466364559,
      "loss": 6.329,
      "mean_token_accuracy": 0.14105381667613984,
      "num_tokens": 3970019.0,
      "step": 2130
    },
    {
      "entropy": 6.581699275970459,
      "epoch": 0.1885376192158248,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998866493583541,
      "loss": 6.3318,
      "mean_token_accuracy": 0.1405398890376091,
      "num_tokens": 3978625.0,
      "step": 2135
    },
    {
      "entropy": 6.564266300201416,
      "epoch": 0.18897915930766512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998856476748509,
      "loss": 6.4541,
      "mean_token_accuracy": 0.13665426075458526,
      "num_tokens": 3989191.0,
      "step": 2140
    },
    {
      "entropy": 6.618397665023804,
      "epoch": 0.18942069939950548,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998846415859656,
      "loss": 6.431,
      "mean_token_accuracy": 0.13899025768041612,
      "num_tokens": 3998558.0,
      "step": 2145
    },
    {
      "entropy": 6.506017112731934,
      "epoch": 0.1898622394913458,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998836310917182,
      "loss": 6.3985,
      "mean_token_accuracy": 0.14203351140022277,
      "num_tokens": 4008318.0,
      "step": 2150
    },
    {
      "entropy": 6.643159866333008,
      "epoch": 0.19030377958318614,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998826161921282,
      "loss": 6.3839,
      "mean_token_accuracy": 0.1404844231903553,
      "num_tokens": 4017133.0,
      "step": 2155
    },
    {
      "entropy": 6.56635799407959,
      "epoch": 0.1907453196750265,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998815968872157,
      "loss": 6.3718,
      "mean_token_accuracy": 0.14540308713912964,
      "num_tokens": 4026289.0,
      "step": 2160
    },
    {
      "entropy": 6.529609298706054,
      "epoch": 0.19118685976686683,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998805731770007,
      "loss": 6.3744,
      "mean_token_accuracy": 0.14073018357157707,
      "num_tokens": 4035209.0,
      "step": 2165
    },
    {
      "entropy": 6.62944221496582,
      "epoch": 0.19162839985870717,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499879545061503,
      "loss": 6.5068,
      "mean_token_accuracy": 0.12898163348436356,
      "num_tokens": 4044498.0,
      "step": 2170
    },
    {
      "entropy": 6.6656512260437015,
      "epoch": 0.1920699399505475,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998785125407432,
      "loss": 6.4006,
      "mean_token_accuracy": 0.14504364654421806,
      "num_tokens": 4054211.0,
      "step": 2175
    },
    {
      "entropy": 6.602165699005127,
      "epoch": 0.19251148004238786,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499877475614741,
      "loss": 6.4231,
      "mean_token_accuracy": 0.13061779513955116,
      "num_tokens": 4064089.0,
      "step": 2180
    },
    {
      "entropy": 6.533041334152221,
      "epoch": 0.1929530201342282,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998764342835169,
      "loss": 6.3658,
      "mean_token_accuracy": 0.14450084567070007,
      "num_tokens": 4073042.0,
      "step": 2185
    },
    {
      "entropy": 6.503344345092773,
      "epoch": 0.19339456022606852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998753885470915,
      "loss": 6.4349,
      "mean_token_accuracy": 0.13570086807012557,
      "num_tokens": 4081667.0,
      "step": 2190
    },
    {
      "entropy": 6.554050397872925,
      "epoch": 0.19383610031790888,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998743384054851,
      "loss": 6.4315,
      "mean_token_accuracy": 0.14233472645282746,
      "num_tokens": 4090598.0,
      "step": 2195
    },
    {
      "entropy": 6.641329669952393,
      "epoch": 0.1942776404097492,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004998732838587183,
      "loss": 6.3621,
      "mean_token_accuracy": 0.14019383788108825,
      "num_tokens": 4100113.0,
      "step": 2200
    },
    {
      "entropy": 6.567934036254883,
      "epoch": 0.19471918050158954,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998722249068118,
      "loss": 6.3488,
      "mean_token_accuracy": 0.14066357761621476,
      "num_tokens": 4108473.0,
      "step": 2205
    },
    {
      "entropy": 6.440048599243164,
      "epoch": 0.19516072059342987,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998711615497863,
      "loss": 6.3735,
      "mean_token_accuracy": 0.14175349250435829,
      "num_tokens": 4118479.0,
      "step": 2210
    },
    {
      "entropy": 6.5755712509155275,
      "epoch": 0.19560226068527023,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998700937876626,
      "loss": 6.326,
      "mean_token_accuracy": 0.1472289428114891,
      "num_tokens": 4127458.0,
      "step": 2215
    },
    {
      "entropy": 6.598430156707764,
      "epoch": 0.19604380077711056,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004998690216204615,
      "loss": 6.4709,
      "mean_token_accuracy": 0.13767264410853386,
      "num_tokens": 4138099.0,
      "step": 2220
    },
    {
      "entropy": 6.552243328094482,
      "epoch": 0.1964853408689509,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998679450482043,
      "loss": 6.335,
      "mean_token_accuracy": 0.14231890067458153,
      "num_tokens": 4147991.0,
      "step": 2225
    },
    {
      "entropy": 6.487472057342529,
      "epoch": 0.19692688096079125,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499866864070912,
      "loss": 6.4582,
      "mean_token_accuracy": 0.1360939234495163,
      "num_tokens": 4157128.0,
      "step": 2230
    },
    {
      "entropy": 6.669093990325928,
      "epoch": 0.19736842105263158,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998657786886056,
      "loss": 6.4883,
      "mean_token_accuracy": 0.13292337208986282,
      "num_tokens": 4168339.0,
      "step": 2235
    },
    {
      "entropy": 6.5326982021331785,
      "epoch": 0.1978099611444719,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998646889013066,
      "loss": 6.3211,
      "mean_token_accuracy": 0.14647760689258577,
      "num_tokens": 4177476.0,
      "step": 2240
    },
    {
      "entropy": 6.605254316329956,
      "epoch": 0.19825150123631224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998635947090362,
      "loss": 6.4406,
      "mean_token_accuracy": 0.1364073134958744,
      "num_tokens": 4187578.0,
      "step": 2245
    },
    {
      "entropy": 6.487747430801392,
      "epoch": 0.1986930413281526,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998624961118158,
      "loss": 6.3272,
      "mean_token_accuracy": 0.14302509576082229,
      "num_tokens": 4196826.0,
      "step": 2250
    },
    {
      "entropy": 6.587798023223877,
      "epoch": 0.19913458141999293,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499861393109667,
      "loss": 6.301,
      "mean_token_accuracy": 0.14366807490587236,
      "num_tokens": 4206085.0,
      "step": 2255
    },
    {
      "entropy": 6.432883596420288,
      "epoch": 0.19957612151183327,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998602857026114,
      "loss": 6.3367,
      "mean_token_accuracy": 0.1540941886603832,
      "num_tokens": 4215375.0,
      "step": 2260
    },
    {
      "entropy": 6.59939227104187,
      "epoch": 0.20001766160367362,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998591738906708,
      "loss": 6.3942,
      "mean_token_accuracy": 0.13624251410365104,
      "num_tokens": 4224809.0,
      "step": 2265
    },
    {
      "entropy": 6.535125207901001,
      "epoch": 0.20045920169551396,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998580576738668,
      "loss": 6.3869,
      "mean_token_accuracy": 0.1412413999438286,
      "num_tokens": 4234048.0,
      "step": 2270
    },
    {
      "entropy": 6.533976364135742,
      "epoch": 0.2009007417873543,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998569370522213,
      "loss": 6.4072,
      "mean_token_accuracy": 0.14488251730799676,
      "num_tokens": 4242622.0,
      "step": 2275
    },
    {
      "entropy": 6.589724588394165,
      "epoch": 0.20134228187919462,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998558120257563,
      "loss": 6.3473,
      "mean_token_accuracy": 0.14969526827335358,
      "num_tokens": 4251833.0,
      "step": 2280
    },
    {
      "entropy": 6.640519809722901,
      "epoch": 0.20178382197103498,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998546825944938,
      "loss": 6.4479,
      "mean_token_accuracy": 0.13527315109968185,
      "num_tokens": 4260769.0,
      "step": 2285
    },
    {
      "entropy": 6.484637928009033,
      "epoch": 0.2022253620628753,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499853548758456,
      "loss": 6.4441,
      "mean_token_accuracy": 0.14147843569517135,
      "num_tokens": 4270805.0,
      "step": 2290
    },
    {
      "entropy": 6.580462312698364,
      "epoch": 0.20266690215471564,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499852410517665,
      "loss": 6.4282,
      "mean_token_accuracy": 0.13979952931404113,
      "num_tokens": 4280860.0,
      "step": 2295
    },
    {
      "entropy": 6.549244785308838,
      "epoch": 0.203108442246556,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998512678721431,
      "loss": 6.4135,
      "mean_token_accuracy": 0.14154525697231293,
      "num_tokens": 4291283.0,
      "step": 2300
    },
    {
      "entropy": 6.6056934833526615,
      "epoch": 0.20354998233839633,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499850120821913,
      "loss": 6.3769,
      "mean_token_accuracy": 0.13967125117778778,
      "num_tokens": 4300920.0,
      "step": 2305
    },
    {
      "entropy": 6.631320190429688,
      "epoch": 0.20399152243023666,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998489693669967,
      "loss": 6.3364,
      "mean_token_accuracy": 0.1530125007033348,
      "num_tokens": 4311056.0,
      "step": 2310
    },
    {
      "entropy": 6.41552734375,
      "epoch": 0.204433062522077,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499847813507417,
      "loss": 6.2605,
      "mean_token_accuracy": 0.14134440571069717,
      "num_tokens": 4320013.0,
      "step": 2315
    },
    {
      "entropy": 6.52802586555481,
      "epoch": 0.20487460261391735,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998466532431966,
      "loss": 6.2845,
      "mean_token_accuracy": 0.1438765361905098,
      "num_tokens": 4328243.0,
      "step": 2320
    },
    {
      "entropy": 6.505352210998535,
      "epoch": 0.20531614270575768,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998454885743581,
      "loss": 6.3987,
      "mean_token_accuracy": 0.1398943543434143,
      "num_tokens": 4337126.0,
      "step": 2325
    },
    {
      "entropy": 6.534863376617432,
      "epoch": 0.205757682797598,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998443195009242,
      "loss": 6.3298,
      "mean_token_accuracy": 0.14415593892335893,
      "num_tokens": 4347008.0,
      "step": 2330
    },
    {
      "entropy": 6.526832628250122,
      "epoch": 0.20619922288943837,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998431460229182,
      "loss": 6.3074,
      "mean_token_accuracy": 0.14151543527841567,
      "num_tokens": 4355979.0,
      "step": 2335
    },
    {
      "entropy": 6.50370044708252,
      "epoch": 0.2066407629812787,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998419681403627,
      "loss": 6.3397,
      "mean_token_accuracy": 0.13715255707502366,
      "num_tokens": 4364843.0,
      "step": 2340
    },
    {
      "entropy": 6.5381780624389645,
      "epoch": 0.20708230307311903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998407858532809,
      "loss": 6.399,
      "mean_token_accuracy": 0.1404947891831398,
      "num_tokens": 4373951.0,
      "step": 2345
    },
    {
      "entropy": 6.539831066131592,
      "epoch": 0.20752384316495937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499839599161696,
      "loss": 6.3743,
      "mean_token_accuracy": 0.1348615162074566,
      "num_tokens": 4382672.0,
      "step": 2350
    },
    {
      "entropy": 6.503879594802856,
      "epoch": 0.20796538325679972,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998384080656314,
      "loss": 6.323,
      "mean_token_accuracy": 0.1376982130110264,
      "num_tokens": 4392113.0,
      "step": 2355
    },
    {
      "entropy": 6.495373058319092,
      "epoch": 0.20840692334864006,
      "grad_norm": 1.0,
      "learning_rate": 0.00049983721256511,
      "loss": 6.3301,
      "mean_token_accuracy": 0.14757516384124755,
      "num_tokens": 4401587.0,
      "step": 2360
    },
    {
      "entropy": 6.432989120483398,
      "epoch": 0.2088484634404804,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998360126601556,
      "loss": 6.2641,
      "mean_token_accuracy": 0.14713913276791574,
      "num_tokens": 4410304.0,
      "step": 2365
    },
    {
      "entropy": 6.4272682666778564,
      "epoch": 0.20929000353232075,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998348083507916,
      "loss": 6.386,
      "mean_token_accuracy": 0.1397126257419586,
      "num_tokens": 4419832.0,
      "step": 2370
    },
    {
      "entropy": 6.525888919830322,
      "epoch": 0.20973154362416108,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998335996370416,
      "loss": 6.2407,
      "mean_token_accuracy": 0.15095002949237823,
      "num_tokens": 4428237.0,
      "step": 2375
    },
    {
      "entropy": 6.497446823120117,
      "epoch": 0.2101730837160014,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998323865189291,
      "loss": 6.3849,
      "mean_token_accuracy": 0.13744312450289725,
      "num_tokens": 4437325.0,
      "step": 2380
    },
    {
      "entropy": 6.524771118164063,
      "epoch": 0.21061462380784174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998311689964781,
      "loss": 6.2568,
      "mean_token_accuracy": 0.15043262839317323,
      "num_tokens": 4446745.0,
      "step": 2385
    },
    {
      "entropy": 6.476553869247437,
      "epoch": 0.2110561638996821,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998299470697125,
      "loss": 6.3391,
      "mean_token_accuracy": 0.14306224137544632,
      "num_tokens": 4455947.0,
      "step": 2390
    },
    {
      "entropy": 6.497906303405761,
      "epoch": 0.21149770399152243,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998287207386559,
      "loss": 6.3061,
      "mean_token_accuracy": 0.1393039345741272,
      "num_tokens": 4465300.0,
      "step": 2395
    },
    {
      "entropy": 6.564029312133789,
      "epoch": 0.21193924408336276,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998274900033326,
      "loss": 6.3067,
      "mean_token_accuracy": 0.1415036529302597,
      "num_tokens": 4474470.0,
      "step": 2400
    },
    {
      "entropy": 6.402004289627075,
      "epoch": 0.21238078417520312,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998262548637667,
      "loss": 6.3252,
      "mean_token_accuracy": 0.14604315310716628,
      "num_tokens": 4483991.0,
      "step": 2405
    },
    {
      "entropy": 6.517073488235473,
      "epoch": 0.21282232426704345,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998250153199822,
      "loss": 6.3753,
      "mean_token_accuracy": 0.14476549327373506,
      "num_tokens": 4493963.0,
      "step": 2410
    },
    {
      "entropy": 6.551792812347412,
      "epoch": 0.21326386435888378,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998237713720036,
      "loss": 6.3964,
      "mean_token_accuracy": 0.14275107234716417,
      "num_tokens": 4504310.0,
      "step": 2415
    },
    {
      "entropy": 6.4785685539245605,
      "epoch": 0.2137054044507241,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998225230198552,
      "loss": 6.2939,
      "mean_token_accuracy": 0.1452530562877655,
      "num_tokens": 4513217.0,
      "step": 2420
    },
    {
      "entropy": 6.5528992176055905,
      "epoch": 0.21414694454256447,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998212702635614,
      "loss": 6.365,
      "mean_token_accuracy": 0.13781265765428544,
      "num_tokens": 4521746.0,
      "step": 2425
    },
    {
      "entropy": 6.4305943012237545,
      "epoch": 0.2145884846344048,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998200131031469,
      "loss": 6.2658,
      "mean_token_accuracy": 0.14664409756660463,
      "num_tokens": 4531244.0,
      "step": 2430
    },
    {
      "entropy": 6.574305582046509,
      "epoch": 0.21503002472624513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998187515386361,
      "loss": 6.2638,
      "mean_token_accuracy": 0.14604247286915778,
      "num_tokens": 4540065.0,
      "step": 2435
    },
    {
      "entropy": 6.408740425109864,
      "epoch": 0.2154715648180855,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998174855700538,
      "loss": 6.344,
      "mean_token_accuracy": 0.1382213518023491,
      "num_tokens": 4549691.0,
      "step": 2440
    },
    {
      "entropy": 6.517018890380859,
      "epoch": 0.21591310490992582,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998162151974248,
      "loss": 6.2199,
      "mean_token_accuracy": 0.14915743321180344,
      "num_tokens": 4559311.0,
      "step": 2445
    },
    {
      "entropy": 6.559193229675293,
      "epoch": 0.21635464500176615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499814940420774,
      "loss": 6.3878,
      "mean_token_accuracy": 0.14189895316958429,
      "num_tokens": 4569243.0,
      "step": 2450
    },
    {
      "entropy": 6.393787431716919,
      "epoch": 0.21679618509360649,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998136612401266,
      "loss": 6.326,
      "mean_token_accuracy": 0.1409069925546646,
      "num_tokens": 4578898.0,
      "step": 2455
    },
    {
      "entropy": 6.48945894241333,
      "epoch": 0.21723772518544684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998123776555071,
      "loss": 6.3673,
      "mean_token_accuracy": 0.143455608189106,
      "num_tokens": 4588443.0,
      "step": 2460
    },
    {
      "entropy": 6.497863721847534,
      "epoch": 0.21767926527728718,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998110896669412,
      "loss": 6.3381,
      "mean_token_accuracy": 0.13771693482995034,
      "num_tokens": 4599333.0,
      "step": 2465
    },
    {
      "entropy": 6.558657217025757,
      "epoch": 0.2181208053691275,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998097972744539,
      "loss": 6.3569,
      "mean_token_accuracy": 0.14392640367150306,
      "num_tokens": 4608287.0,
      "step": 2470
    },
    {
      "entropy": 6.453986263275146,
      "epoch": 0.21856234546096787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998085004780705,
      "loss": 6.2777,
      "mean_token_accuracy": 0.14240417033433914,
      "num_tokens": 4616849.0,
      "step": 2475
    },
    {
      "entropy": 6.480883169174194,
      "epoch": 0.2190038855528082,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998071992778164,
      "loss": 6.2876,
      "mean_token_accuracy": 0.14548520371317863,
      "num_tokens": 4625830.0,
      "step": 2480
    },
    {
      "entropy": 6.467753934860229,
      "epoch": 0.21944542564464853,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499805893673717,
      "loss": 6.3282,
      "mean_token_accuracy": 0.14801275581121445,
      "num_tokens": 4635404.0,
      "step": 2485
    },
    {
      "entropy": 6.565742492675781,
      "epoch": 0.21988696573648886,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998045836657982,
      "loss": 6.3555,
      "mean_token_accuracy": 0.14250202104449272,
      "num_tokens": 4644422.0,
      "step": 2490
    },
    {
      "entropy": 6.506637191772461,
      "epoch": 0.22032850582832922,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998032692540853,
      "loss": 6.3906,
      "mean_token_accuracy": 0.13982019126415252,
      "num_tokens": 4653641.0,
      "step": 2495
    },
    {
      "entropy": 6.542034721374511,
      "epoch": 0.22077004592016955,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998019504386044,
      "loss": 6.3289,
      "mean_token_accuracy": 0.143229641020298,
      "num_tokens": 4662999.0,
      "step": 2500
    },
    {
      "entropy": 6.514064455032349,
      "epoch": 0.22121158601200988,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998006272193809,
      "loss": 6.3434,
      "mean_token_accuracy": 0.1452925592660904,
      "num_tokens": 4672587.0,
      "step": 2505
    },
    {
      "entropy": 6.424850225448608,
      "epoch": 0.22165312610385024,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997992995964412,
      "loss": 6.2687,
      "mean_token_accuracy": 0.14280916899442672,
      "num_tokens": 4682504.0,
      "step": 2510
    },
    {
      "entropy": 6.46969838142395,
      "epoch": 0.22209466619569057,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997979675698109,
      "loss": 6.3981,
      "mean_token_accuracy": 0.14228205978870392,
      "num_tokens": 4691728.0,
      "step": 2515
    },
    {
      "entropy": 6.499358606338501,
      "epoch": 0.2225362062875309,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997966311395164,
      "loss": 6.2038,
      "mean_token_accuracy": 0.15347035229206085,
      "num_tokens": 4701169.0,
      "step": 2520
    },
    {
      "entropy": 6.416709661483765,
      "epoch": 0.22297774637937123,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997952903055836,
      "loss": 6.225,
      "mean_token_accuracy": 0.1492363281548023,
      "num_tokens": 4711050.0,
      "step": 2525
    },
    {
      "entropy": 6.540422153472901,
      "epoch": 0.2234192864712116,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499793945068039,
      "loss": 6.3515,
      "mean_token_accuracy": 0.14371454566717148,
      "num_tokens": 4720520.0,
      "step": 2530
    },
    {
      "entropy": 6.360270595550537,
      "epoch": 0.22386082656305192,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997925954269088,
      "loss": 6.2195,
      "mean_token_accuracy": 0.1561760887503624,
      "num_tokens": 4729446.0,
      "step": 2535
    },
    {
      "entropy": 6.379441928863526,
      "epoch": 0.22430236665489225,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997912413822196,
      "loss": 6.3196,
      "mean_token_accuracy": 0.14638870134949683,
      "num_tokens": 4739294.0,
      "step": 2540
    },
    {
      "entropy": 6.594422245025635,
      "epoch": 0.2247439067467326,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997898829339979,
      "loss": 6.2911,
      "mean_token_accuracy": 0.14837272465229034,
      "num_tokens": 4747764.0,
      "step": 2545
    },
    {
      "entropy": 6.369044256210327,
      "epoch": 0.22518544683857294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049978852008227,
      "loss": 6.2216,
      "mean_token_accuracy": 0.1513258472084999,
      "num_tokens": 4756647.0,
      "step": 2550
    },
    {
      "entropy": 6.45693073272705,
      "epoch": 0.22562698693041328,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499787152827063,
      "loss": 6.2796,
      "mean_token_accuracy": 0.14629189819097518,
      "num_tokens": 4765907.0,
      "step": 2555
    },
    {
      "entropy": 6.52665147781372,
      "epoch": 0.2260685270222536,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997857811684035,
      "loss": 6.3609,
      "mean_token_accuracy": 0.14549251645803452,
      "num_tokens": 4774670.0,
      "step": 2560
    },
    {
      "entropy": 6.445423984527588,
      "epoch": 0.22651006711409397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997844051063183,
      "loss": 6.2632,
      "mean_token_accuracy": 0.15502836257219316,
      "num_tokens": 4783662.0,
      "step": 2565
    },
    {
      "entropy": 6.455739736557007,
      "epoch": 0.2269516072059343,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997830246408346,
      "loss": 6.2772,
      "mean_token_accuracy": 0.14837741404771804,
      "num_tokens": 4794352.0,
      "step": 2570
    },
    {
      "entropy": 6.368543529510498,
      "epoch": 0.22739314729777463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997816397719791,
      "loss": 6.2842,
      "mean_token_accuracy": 0.14470751956105232,
      "num_tokens": 4805185.0,
      "step": 2575
    },
    {
      "entropy": 6.49798994064331,
      "epoch": 0.227834687389615,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997802504997792,
      "loss": 6.2227,
      "mean_token_accuracy": 0.1516940973699093,
      "num_tokens": 4814178.0,
      "step": 2580
    },
    {
      "entropy": 6.3884289264678955,
      "epoch": 0.22827622748145532,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997788568242621,
      "loss": 6.2732,
      "mean_token_accuracy": 0.15293048024177552,
      "num_tokens": 4822634.0,
      "step": 2585
    },
    {
      "entropy": 6.4718101024627686,
      "epoch": 0.22871776757329565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499777458745455,
      "loss": 6.2981,
      "mean_token_accuracy": 0.14538061991333961,
      "num_tokens": 4831892.0,
      "step": 2590
    },
    {
      "entropy": 6.53868556022644,
      "epoch": 0.22915930766513598,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997760562633853,
      "loss": 6.2461,
      "mean_token_accuracy": 0.14104571789503098,
      "num_tokens": 4840585.0,
      "step": 2595
    },
    {
      "entropy": 6.3369544506072994,
      "epoch": 0.22960084775697634,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997746493780804,
      "loss": 6.2662,
      "mean_token_accuracy": 0.14956960678100586,
      "num_tokens": 4849680.0,
      "step": 2600
    },
    {
      "entropy": 6.402548980712891,
      "epoch": 0.23004238784881667,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499773238089568,
      "loss": 6.1946,
      "mean_token_accuracy": 0.15193643122911454,
      "num_tokens": 4858935.0,
      "step": 2605
    },
    {
      "entropy": 6.451043272018433,
      "epoch": 0.230483927940657,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997718223978758,
      "loss": 6.2793,
      "mean_token_accuracy": 0.15012711733579637,
      "num_tokens": 4868612.0,
      "step": 2610
    },
    {
      "entropy": 6.471085453033448,
      "epoch": 0.23092546803249736,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997704023030315,
      "loss": 6.3092,
      "mean_token_accuracy": 0.14472778290510177,
      "num_tokens": 4876923.0,
      "step": 2615
    },
    {
      "entropy": 6.488173103332519,
      "epoch": 0.2313670081243377,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997689778050627,
      "loss": 6.3912,
      "mean_token_accuracy": 0.1424463249742985,
      "num_tokens": 4886683.0,
      "step": 2620
    },
    {
      "entropy": 6.476648426055908,
      "epoch": 0.23180854821617802,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997675489039975,
      "loss": 6.2493,
      "mean_token_accuracy": 0.14954889714717864,
      "num_tokens": 4895877.0,
      "step": 2625
    },
    {
      "entropy": 6.476872062683105,
      "epoch": 0.23225008830801835,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997661155998638,
      "loss": 6.3645,
      "mean_token_accuracy": 0.1469063326716423,
      "num_tokens": 4904929.0,
      "step": 2630
    },
    {
      "entropy": 6.5085193634033205,
      "epoch": 0.2326916283998587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997646778926898,
      "loss": 6.2951,
      "mean_token_accuracy": 0.14097197651863097,
      "num_tokens": 4913157.0,
      "step": 2635
    },
    {
      "entropy": 6.490143060684204,
      "epoch": 0.23313316849169904,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997632357825035,
      "loss": 6.3681,
      "mean_token_accuracy": 0.13852858543395996,
      "num_tokens": 4923791.0,
      "step": 2640
    },
    {
      "entropy": 6.4098340511322025,
      "epoch": 0.23357470858353938,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997617892693333,
      "loss": 6.2606,
      "mean_token_accuracy": 0.1469244882464409,
      "num_tokens": 4933184.0,
      "step": 2645
    },
    {
      "entropy": 6.462233257293701,
      "epoch": 0.23401624867537973,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997603383532075,
      "loss": 6.2178,
      "mean_token_accuracy": 0.15098182260990142,
      "num_tokens": 4941833.0,
      "step": 2650
    },
    {
      "entropy": 6.405755138397216,
      "epoch": 0.23445778876722007,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997588830341545,
      "loss": 6.1816,
      "mean_token_accuracy": 0.15357830375432968,
      "num_tokens": 4950942.0,
      "step": 2655
    },
    {
      "entropy": 6.4815491199493405,
      "epoch": 0.2348993288590604,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997574233122028,
      "loss": 6.3526,
      "mean_token_accuracy": 0.1432811886072159,
      "num_tokens": 4962005.0,
      "step": 2660
    },
    {
      "entropy": 6.4480256080627445,
      "epoch": 0.23534086895090076,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997559591873809,
      "loss": 6.2118,
      "mean_token_accuracy": 0.15217095017433166,
      "num_tokens": 4970965.0,
      "step": 2665
    },
    {
      "entropy": 6.428168153762817,
      "epoch": 0.2357824090427411,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997544906597178,
      "loss": 6.2401,
      "mean_token_accuracy": 0.14764407500624657,
      "num_tokens": 4980460.0,
      "step": 2670
    },
    {
      "entropy": 6.38992223739624,
      "epoch": 0.23622394913458142,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997530177292418,
      "loss": 6.1835,
      "mean_token_accuracy": 0.1445123538374901,
      "num_tokens": 4989343.0,
      "step": 2675
    },
    {
      "entropy": 6.2847757816314695,
      "epoch": 0.23666548922642175,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997515403959823,
      "loss": 6.2281,
      "mean_token_accuracy": 0.14587305262684822,
      "num_tokens": 4998881.0,
      "step": 2680
    },
    {
      "entropy": 6.461613607406616,
      "epoch": 0.2371070293182621,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997500586599677,
      "loss": 6.292,
      "mean_token_accuracy": 0.14370835274457933,
      "num_tokens": 5008368.0,
      "step": 2685
    },
    {
      "entropy": 6.479917383193969,
      "epoch": 0.23754856941010244,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997485725212274,
      "loss": 6.3201,
      "mean_token_accuracy": 0.15082732439041138,
      "num_tokens": 5017904.0,
      "step": 2690
    },
    {
      "entropy": 6.5032206058502195,
      "epoch": 0.23799010950194277,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997470819797903,
      "loss": 6.2612,
      "mean_token_accuracy": 0.15601224899291993,
      "num_tokens": 5026836.0,
      "step": 2695
    },
    {
      "entropy": 6.387510490417481,
      "epoch": 0.23843164959378313,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997455870356857,
      "loss": 6.2253,
      "mean_token_accuracy": 0.14688334614038467,
      "num_tokens": 5035621.0,
      "step": 2700
    },
    {
      "entropy": 6.4787116050720215,
      "epoch": 0.23887318968562346,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997440876889429,
      "loss": 6.2895,
      "mean_token_accuracy": 0.1482943207025528,
      "num_tokens": 5044076.0,
      "step": 2705
    },
    {
      "entropy": 6.492949533462524,
      "epoch": 0.2393147297774638,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997425839395913,
      "loss": 6.3029,
      "mean_token_accuracy": 0.14274363666772844,
      "num_tokens": 5053931.0,
      "step": 2710
    },
    {
      "entropy": 6.3609747886657715,
      "epoch": 0.23975626986930412,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997410757876602,
      "loss": 6.2358,
      "mean_token_accuracy": 0.15066074430942536,
      "num_tokens": 5063969.0,
      "step": 2715
    },
    {
      "entropy": 6.469989538192749,
      "epoch": 0.24019780996114448,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997395632331793,
      "loss": 6.3041,
      "mean_token_accuracy": 0.1426872693002224,
      "num_tokens": 5073253.0,
      "step": 2720
    },
    {
      "entropy": 6.411794424057007,
      "epoch": 0.2406393500529848,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997380462761781,
      "loss": 6.2456,
      "mean_token_accuracy": 0.1446852631866932,
      "num_tokens": 5082673.0,
      "step": 2725
    },
    {
      "entropy": 6.405149459838867,
      "epoch": 0.24108089014482514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997365249166864,
      "loss": 6.2601,
      "mean_token_accuracy": 0.15146871134638787,
      "num_tokens": 5092040.0,
      "step": 2730
    },
    {
      "entropy": 6.448270606994629,
      "epoch": 0.2415224302366655,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997349991547342,
      "loss": 6.2359,
      "mean_token_accuracy": 0.14747195094823837,
      "num_tokens": 5101942.0,
      "step": 2735
    },
    {
      "entropy": 6.3382261276245115,
      "epoch": 0.24196397032850583,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997334689903509,
      "loss": 6.2302,
      "mean_token_accuracy": 0.15405145585536956,
      "num_tokens": 5111081.0,
      "step": 2740
    },
    {
      "entropy": 6.4858105182647705,
      "epoch": 0.24240551042034617,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997319344235668,
      "loss": 6.2757,
      "mean_token_accuracy": 0.1433402493596077,
      "num_tokens": 5120250.0,
      "step": 2745
    },
    {
      "entropy": 6.476590394973755,
      "epoch": 0.2428470505121865,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499730395454412,
      "loss": 6.3262,
      "mean_token_accuracy": 0.13439106568694115,
      "num_tokens": 5130090.0,
      "step": 2750
    },
    {
      "entropy": 6.378527879714966,
      "epoch": 0.24328859060402686,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997288520829166,
      "loss": 6.2372,
      "mean_token_accuracy": 0.14988442063331603,
      "num_tokens": 5138680.0,
      "step": 2755
    },
    {
      "entropy": 6.441170692443848,
      "epoch": 0.2437301306958672,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997273043091107,
      "loss": 6.304,
      "mean_token_accuracy": 0.1427198089659214,
      "num_tokens": 5150025.0,
      "step": 2760
    },
    {
      "entropy": 6.356282091140747,
      "epoch": 0.24417167078770752,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997257521330248,
      "loss": 6.1376,
      "mean_token_accuracy": 0.15772749930620195,
      "num_tokens": 5159159.0,
      "step": 2765
    },
    {
      "entropy": 6.350546789169312,
      "epoch": 0.24461321087954788,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997241955546892,
      "loss": 6.3025,
      "mean_token_accuracy": 0.1383981317281723,
      "num_tokens": 5168584.0,
      "step": 2770
    },
    {
      "entropy": 6.4935712814331055,
      "epoch": 0.2450547509713882,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997226345741343,
      "loss": 6.2395,
      "mean_token_accuracy": 0.14813449531793593,
      "num_tokens": 5177804.0,
      "step": 2775
    },
    {
      "entropy": 6.330422210693359,
      "epoch": 0.24549629106322854,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499721069191391,
      "loss": 6.1723,
      "mean_token_accuracy": 0.15173402577638626,
      "num_tokens": 5187022.0,
      "step": 2780
    },
    {
      "entropy": 6.383345079421997,
      "epoch": 0.24593783115506887,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997194994064896,
      "loss": 6.1975,
      "mean_token_accuracy": 0.15482074320316314,
      "num_tokens": 5195882.0,
      "step": 2785
    },
    {
      "entropy": 6.353169202804565,
      "epoch": 0.24637937124690923,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499717925219461,
      "loss": 6.1664,
      "mean_token_accuracy": 0.14942888617515565,
      "num_tokens": 5204401.0,
      "step": 2790
    },
    {
      "entropy": 6.470758771896362,
      "epoch": 0.24682091133874956,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997163466303362,
      "loss": 6.1492,
      "mean_token_accuracy": 0.15647954791784285,
      "num_tokens": 5214262.0,
      "step": 2795
    },
    {
      "entropy": 6.308819770812988,
      "epoch": 0.2472624514305899,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499714763639146,
      "loss": 6.2008,
      "mean_token_accuracy": 0.14968698844313622,
      "num_tokens": 5223871.0,
      "step": 2800
    },
    {
      "entropy": 6.387062311172485,
      "epoch": 0.24770399152243025,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997131762459211,
      "loss": 6.2141,
      "mean_token_accuracy": 0.15620359778404236,
      "num_tokens": 5233395.0,
      "step": 2805
    },
    {
      "entropy": 6.389224910736084,
      "epoch": 0.24814553161427058,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997115844506932,
      "loss": 6.1637,
      "mean_token_accuracy": 0.15392670035362244,
      "num_tokens": 5243184.0,
      "step": 2810
    },
    {
      "entropy": 6.247744512557984,
      "epoch": 0.2485870717061109,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997099882534929,
      "loss": 6.0597,
      "mean_token_accuracy": 0.16176963746547698,
      "num_tokens": 5252851.0,
      "step": 2815
    },
    {
      "entropy": 6.453801250457763,
      "epoch": 0.24902861179795124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997083876543519,
      "loss": 6.2417,
      "mean_token_accuracy": 0.14792601466178895,
      "num_tokens": 5261870.0,
      "step": 2820
    },
    {
      "entropy": 6.370384979248047,
      "epoch": 0.2494701518897916,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997067826533014,
      "loss": 6.1957,
      "mean_token_accuracy": 0.1431150496006012,
      "num_tokens": 5271639.0,
      "step": 2825
    },
    {
      "entropy": 6.440698575973511,
      "epoch": 0.24991169198163193,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997051732503726,
      "loss": 6.2035,
      "mean_token_accuracy": 0.14459088742733,
      "num_tokens": 5280746.0,
      "step": 2830
    },
    {
      "entropy": 6.372076845169067,
      "epoch": 0.25035323207347226,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997035594455975,
      "loss": 6.2268,
      "mean_token_accuracy": 0.14702461585402488,
      "num_tokens": 5290201.0,
      "step": 2835
    },
    {
      "entropy": 6.380630159378052,
      "epoch": 0.2507947721653126,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997019412390074,
      "loss": 6.2107,
      "mean_token_accuracy": 0.15259625762701035,
      "num_tokens": 5300336.0,
      "step": 2840
    },
    {
      "entropy": 6.392699670791626,
      "epoch": 0.2512363122571529,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499700318630634,
      "loss": 6.2537,
      "mean_token_accuracy": 0.14918554574251175,
      "num_tokens": 5309174.0,
      "step": 2845
    },
    {
      "entropy": 6.462964773178101,
      "epoch": 0.2516778523489933,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996986916205092,
      "loss": 6.3122,
      "mean_token_accuracy": 0.14475011378526687,
      "num_tokens": 5317702.0,
      "step": 2850
    },
    {
      "entropy": 6.428555202484131,
      "epoch": 0.25211939244083365,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996970602086648,
      "loss": 6.2571,
      "mean_token_accuracy": 0.1455409847199917,
      "num_tokens": 5326838.0,
      "step": 2855
    },
    {
      "entropy": 6.358886051177978,
      "epoch": 0.252560932532674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996954243951327,
      "loss": 6.1463,
      "mean_token_accuracy": 0.15396089106798172,
      "num_tokens": 5335679.0,
      "step": 2860
    },
    {
      "entropy": 6.4643134593963625,
      "epoch": 0.2530024726245143,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996937841799451,
      "loss": 6.2858,
      "mean_token_accuracy": 0.15384045615792274,
      "num_tokens": 5344670.0,
      "step": 2865
    },
    {
      "entropy": 6.345240116119385,
      "epoch": 0.25344401271635464,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996921395631342,
      "loss": 6.1141,
      "mean_token_accuracy": 0.15072092190384864,
      "num_tokens": 5353501.0,
      "step": 2870
    },
    {
      "entropy": 6.391109418869019,
      "epoch": 0.25388555280819497,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499690490544732,
      "loss": 6.2151,
      "mean_token_accuracy": 0.15420024394989013,
      "num_tokens": 5362778.0,
      "step": 2875
    },
    {
      "entropy": 6.355072355270385,
      "epoch": 0.2543270929000353,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996888371247707,
      "loss": 6.187,
      "mean_token_accuracy": 0.15069863200187683,
      "num_tokens": 5371231.0,
      "step": 2880
    },
    {
      "entropy": 6.396897888183593,
      "epoch": 0.2547686329918757,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499687179303283,
      "loss": 6.2352,
      "mean_token_accuracy": 0.14393327459692956,
      "num_tokens": 5380655.0,
      "step": 2885
    },
    {
      "entropy": 6.39494571685791,
      "epoch": 0.255210173083716,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996855170803012,
      "loss": 6.2326,
      "mean_token_accuracy": 0.14647915214300156,
      "num_tokens": 5390269.0,
      "step": 2890
    },
    {
      "entropy": 6.329206228256226,
      "epoch": 0.25565171317555635,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996838504558581,
      "loss": 6.0774,
      "mean_token_accuracy": 0.1588391736149788,
      "num_tokens": 5398348.0,
      "step": 2895
    },
    {
      "entropy": 6.267703342437744,
      "epoch": 0.2560932532673967,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499682179429986,
      "loss": 6.2315,
      "mean_token_accuracy": 0.15087486952543258,
      "num_tokens": 5408294.0,
      "step": 2900
    },
    {
      "entropy": 6.416153049468994,
      "epoch": 0.256534793359237,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996805040027178,
      "loss": 6.2348,
      "mean_token_accuracy": 0.15022216588258744,
      "num_tokens": 5418301.0,
      "step": 2905
    },
    {
      "entropy": 6.394410037994385,
      "epoch": 0.25697633345107734,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996788241740863,
      "loss": 6.3045,
      "mean_token_accuracy": 0.1489267498254776,
      "num_tokens": 5427990.0,
      "step": 2910
    },
    {
      "entropy": 6.474830102920532,
      "epoch": 0.2574178735429177,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996771399441243,
      "loss": 6.2283,
      "mean_token_accuracy": 0.1464369684457779,
      "num_tokens": 5437140.0,
      "step": 2915
    },
    {
      "entropy": 6.340023994445801,
      "epoch": 0.25785941363475806,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996754513128652,
      "loss": 6.1935,
      "mean_token_accuracy": 0.15139320865273476,
      "num_tokens": 5445832.0,
      "step": 2920
    },
    {
      "entropy": 6.317194700241089,
      "epoch": 0.2583009537265984,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996737582803416,
      "loss": 6.2144,
      "mean_token_accuracy": 0.1541209414601326,
      "num_tokens": 5455359.0,
      "step": 2925
    },
    {
      "entropy": 6.433325910568238,
      "epoch": 0.2587424938184387,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996720608465868,
      "loss": 6.2866,
      "mean_token_accuracy": 0.1430448040366173,
      "num_tokens": 5465055.0,
      "step": 2930
    },
    {
      "entropy": 6.408900499343872,
      "epoch": 0.25918403391027905,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996703590116342,
      "loss": 6.2059,
      "mean_token_accuracy": 0.15415119379758835,
      "num_tokens": 5474048.0,
      "step": 2935
    },
    {
      "entropy": 6.290044689178467,
      "epoch": 0.2596255740021194,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996686527755171,
      "loss": 6.0412,
      "mean_token_accuracy": 0.1648414745926857,
      "num_tokens": 5483025.0,
      "step": 2940
    },
    {
      "entropy": 6.291623878479004,
      "epoch": 0.2600671140939597,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996669421382687,
      "loss": 6.1957,
      "mean_token_accuracy": 0.15358467102050782,
      "num_tokens": 5492001.0,
      "step": 2945
    },
    {
      "entropy": 6.4285280227661135,
      "epoch": 0.26050865418580005,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004996652270999228,
      "loss": 6.2239,
      "mean_token_accuracy": 0.14998335987329484,
      "num_tokens": 5502524.0,
      "step": 2950
    },
    {
      "entropy": 6.320462036132812,
      "epoch": 0.26095019427764043,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996635076605128,
      "loss": 6.1768,
      "mean_token_accuracy": 0.1475066915154457,
      "num_tokens": 5511526.0,
      "step": 2955
    },
    {
      "entropy": 6.406644201278686,
      "epoch": 0.26139173436948077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996617838200725,
      "loss": 6.2645,
      "mean_token_accuracy": 0.14729123413562775,
      "num_tokens": 5520577.0,
      "step": 2960
    },
    {
      "entropy": 6.391732931137085,
      "epoch": 0.2618332744613211,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996600555786357,
      "loss": 6.155,
      "mean_token_accuracy": 0.15534588098526,
      "num_tokens": 5529681.0,
      "step": 2965
    },
    {
      "entropy": 6.326655769348145,
      "epoch": 0.26227481455316143,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996583229362362,
      "loss": 6.2055,
      "mean_token_accuracy": 0.14942387789487838,
      "num_tokens": 5538984.0,
      "step": 2970
    },
    {
      "entropy": 6.3987696170806885,
      "epoch": 0.26271635464500176,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996565858929078,
      "loss": 6.2258,
      "mean_token_accuracy": 0.1435547314584255,
      "num_tokens": 5548593.0,
      "step": 2975
    },
    {
      "entropy": 6.368779802322388,
      "epoch": 0.2631578947368421,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996548444486847,
      "loss": 6.1997,
      "mean_token_accuracy": 0.15220200568437575,
      "num_tokens": 5557584.0,
      "step": 2980
    },
    {
      "entropy": 6.44010329246521,
      "epoch": 0.2635994348286824,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996530986036008,
      "loss": 6.2569,
      "mean_token_accuracy": 0.14837291613221168,
      "num_tokens": 5567438.0,
      "step": 2985
    },
    {
      "entropy": 6.328577947616577,
      "epoch": 0.2640409749205228,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996513483576907,
      "loss": 6.2752,
      "mean_token_accuracy": 0.1468944802880287,
      "num_tokens": 5577138.0,
      "step": 2990
    },
    {
      "entropy": 6.488377714157105,
      "epoch": 0.26448251501236314,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996495937109884,
      "loss": 6.2327,
      "mean_token_accuracy": 0.1452515721321106,
      "num_tokens": 5586086.0,
      "step": 2995
    },
    {
      "entropy": 6.347766876220703,
      "epoch": 0.26492405510420347,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996478346635283,
      "loss": 6.2034,
      "mean_token_accuracy": 0.1538815215229988,
      "num_tokens": 5594991.0,
      "step": 3000
    },
    {
      "epoch": 0.26492405510420347,
      "eval_entropy": 6.11064261509905,
      "eval_loss": 6.222682952880859,
      "eval_mean_token_accuracy": 0.15619063689225773,
      "eval_num_tokens": 5594991.0,
      "eval_runtime": 20.7082,
      "eval_samples_per_second": 1705.363,
      "eval_steps_per_second": 213.201,
      "step": 3000
    },
    {
      "entropy": 6.42329592704773,
      "epoch": 0.2653655951960438,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996460712153448,
      "loss": 6.2467,
      "mean_token_accuracy": 0.1503360629081726,
      "num_tokens": 5605147.0,
      "step": 3005
    },
    {
      "entropy": 6.324493503570556,
      "epoch": 0.26580713528788413,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996443033664726,
      "loss": 6.1945,
      "mean_token_accuracy": 0.15000912100076674,
      "num_tokens": 5615097.0,
      "step": 3010
    },
    {
      "entropy": 6.3545369625091555,
      "epoch": 0.26624867537972446,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996425311169463,
      "loss": 6.1044,
      "mean_token_accuracy": 0.1567700281739235,
      "num_tokens": 5624026.0,
      "step": 3015
    },
    {
      "entropy": 6.292732095718383,
      "epoch": 0.2666902154715648,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996407544668005,
      "loss": 6.1321,
      "mean_token_accuracy": 0.14880604222416877,
      "num_tokens": 5634029.0,
      "step": 3020
    },
    {
      "entropy": 6.4811278820037845,
      "epoch": 0.2671317555634052,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996389734160701,
      "loss": 6.1977,
      "mean_token_accuracy": 0.15609973818063735,
      "num_tokens": 5642423.0,
      "step": 3025
    },
    {
      "entropy": 6.3460346221923825,
      "epoch": 0.2675732956552455,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049963718796479,
      "loss": 6.1707,
      "mean_token_accuracy": 0.15581064224243163,
      "num_tokens": 5652001.0,
      "step": 3030
    },
    {
      "entropy": 6.269883537292481,
      "epoch": 0.26801483574708584,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996353981129952,
      "loss": 6.0707,
      "mean_token_accuracy": 0.1529800459742546,
      "num_tokens": 5660205.0,
      "step": 3035
    },
    {
      "entropy": 6.290252685546875,
      "epoch": 0.2684563758389262,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996336038607206,
      "loss": 6.1106,
      "mean_token_accuracy": 0.15594022423028947,
      "num_tokens": 5669223.0,
      "step": 3040
    },
    {
      "entropy": 6.309386587142944,
      "epoch": 0.2688979159307665,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996318052080015,
      "loss": 6.1594,
      "mean_token_accuracy": 0.15721056163311004,
      "num_tokens": 5677799.0,
      "step": 3045
    },
    {
      "entropy": 6.3498008251190186,
      "epoch": 0.26933945602260684,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996300021548731,
      "loss": 6.1651,
      "mean_token_accuracy": 0.15220175087451934,
      "num_tokens": 5686864.0,
      "step": 3050
    },
    {
      "entropy": 6.3071695327758786,
      "epoch": 0.26978099611444717,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996281947013707,
      "loss": 6.1839,
      "mean_token_accuracy": 0.15576347708702087,
      "num_tokens": 5697236.0,
      "step": 3055
    },
    {
      "entropy": 6.307339096069336,
      "epoch": 0.27022253620628756,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996263828475298,
      "loss": 6.2356,
      "mean_token_accuracy": 0.14805331826210022,
      "num_tokens": 5706636.0,
      "step": 3060
    },
    {
      "entropy": 6.417330169677735,
      "epoch": 0.2706640762981279,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996245665933857,
      "loss": 6.157,
      "mean_token_accuracy": 0.15448612868785858,
      "num_tokens": 5716698.0,
      "step": 3065
    },
    {
      "entropy": 6.428673553466797,
      "epoch": 0.2711056163899682,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996227459389741,
      "loss": 6.2203,
      "mean_token_accuracy": 0.14477046579122543,
      "num_tokens": 5726594.0,
      "step": 3070
    },
    {
      "entropy": 6.308225297927857,
      "epoch": 0.27154715648180855,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996209208843307,
      "loss": 6.1809,
      "mean_token_accuracy": 0.15280015468597413,
      "num_tokens": 5734715.0,
      "step": 3075
    },
    {
      "entropy": 6.3414054870605465,
      "epoch": 0.2719886965736489,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996190914294912,
      "loss": 6.0912,
      "mean_token_accuracy": 0.15559991002082824,
      "num_tokens": 5743714.0,
      "step": 3080
    },
    {
      "entropy": 6.373754549026489,
      "epoch": 0.2724302366654892,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996172575744914,
      "loss": 6.1923,
      "mean_token_accuracy": 0.15226987898349761,
      "num_tokens": 5753035.0,
      "step": 3085
    },
    {
      "entropy": 6.2699789047241214,
      "epoch": 0.27287177675732954,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996154193193673,
      "loss": 6.1014,
      "mean_token_accuracy": 0.1601513594388962,
      "num_tokens": 5761676.0,
      "step": 3090
    },
    {
      "entropy": 6.389098072052002,
      "epoch": 0.27331331684916993,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996135766641549,
      "loss": 6.1519,
      "mean_token_accuracy": 0.1581235110759735,
      "num_tokens": 5770004.0,
      "step": 3095
    },
    {
      "entropy": 6.378294181823731,
      "epoch": 0.27375485694101026,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996117296088903,
      "loss": 6.1666,
      "mean_token_accuracy": 0.1551177069544792,
      "num_tokens": 5779494.0,
      "step": 3100
    },
    {
      "entropy": 6.267272090911865,
      "epoch": 0.2741963970328506,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996098781536095,
      "loss": 6.1101,
      "mean_token_accuracy": 0.15625676065683364,
      "num_tokens": 5788624.0,
      "step": 3105
    },
    {
      "entropy": 6.379684734344482,
      "epoch": 0.2746379371246909,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996080222983492,
      "loss": 6.2414,
      "mean_token_accuracy": 0.15033891052007675,
      "num_tokens": 5796595.0,
      "step": 3110
    },
    {
      "entropy": 6.3470635414123535,
      "epoch": 0.27507947721653125,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996061620431453,
      "loss": 6.1808,
      "mean_token_accuracy": 0.15361424684524536,
      "num_tokens": 5805716.0,
      "step": 3115
    },
    {
      "entropy": 6.324425220489502,
      "epoch": 0.2755210173083716,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996042973880344,
      "loss": 6.1589,
      "mean_token_accuracy": 0.14875018000602722,
      "num_tokens": 5815200.0,
      "step": 3120
    },
    {
      "entropy": 6.324033451080322,
      "epoch": 0.2759625574002119,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996024283330532,
      "loss": 6.2225,
      "mean_token_accuracy": 0.14702069014310837,
      "num_tokens": 5825613.0,
      "step": 3125
    },
    {
      "entropy": 6.402164268493652,
      "epoch": 0.2764040974920523,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499600554878238,
      "loss": 6.1436,
      "mean_token_accuracy": 0.15741681009531022,
      "num_tokens": 5835486.0,
      "step": 3130
    },
    {
      "entropy": 6.245097303390503,
      "epoch": 0.27684563758389263,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995986770236258,
      "loss": 6.2164,
      "mean_token_accuracy": 0.14954991564154624,
      "num_tokens": 5844878.0,
      "step": 3135
    },
    {
      "entropy": 6.407863521575928,
      "epoch": 0.27728717767573297,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995967947692533,
      "loss": 6.2361,
      "mean_token_accuracy": 0.14730829298496245,
      "num_tokens": 5855367.0,
      "step": 3140
    },
    {
      "entropy": 6.508746910095215,
      "epoch": 0.2777287177675733,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995949081151571,
      "loss": 6.2742,
      "mean_token_accuracy": 0.1438572347164154,
      "num_tokens": 5864489.0,
      "step": 3145
    },
    {
      "entropy": 6.334603500366211,
      "epoch": 0.2781702578594136,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995930170613746,
      "loss": 6.1829,
      "mean_token_accuracy": 0.14453371241688728,
      "num_tokens": 5874803.0,
      "step": 3150
    },
    {
      "entropy": 6.344500398635864,
      "epoch": 0.27861179795125396,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995911216079425,
      "loss": 6.1838,
      "mean_token_accuracy": 0.15842330157756807,
      "num_tokens": 5884355.0,
      "step": 3155
    },
    {
      "entropy": 6.303672170639038,
      "epoch": 0.2790533380430943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995892217548981,
      "loss": 6.1252,
      "mean_token_accuracy": 0.15792049318552018,
      "num_tokens": 5893152.0,
      "step": 3160
    },
    {
      "entropy": 6.400051259994507,
      "epoch": 0.2794948781349347,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995873175022786,
      "loss": 6.2127,
      "mean_token_accuracy": 0.1483101636171341,
      "num_tokens": 5903365.0,
      "step": 3165
    },
    {
      "entropy": 6.312482595443726,
      "epoch": 0.279936418226775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995854088501213,
      "loss": 6.1533,
      "mean_token_accuracy": 0.1574320912361145,
      "num_tokens": 5911891.0,
      "step": 3170
    },
    {
      "entropy": 6.323812055587768,
      "epoch": 0.28037795831861534,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995834957984634,
      "loss": 6.0749,
      "mean_token_accuracy": 0.16025345772504807,
      "num_tokens": 5920964.0,
      "step": 3175
    },
    {
      "entropy": 6.365602302551269,
      "epoch": 0.28081949841045567,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995815783473428,
      "loss": 6.1625,
      "mean_token_accuracy": 0.15167084857821464,
      "num_tokens": 5930912.0,
      "step": 3180
    },
    {
      "entropy": 6.2817144870758055,
      "epoch": 0.281261038502296,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995796564967967,
      "loss": 6.1651,
      "mean_token_accuracy": 0.14691652208566666,
      "num_tokens": 5940489.0,
      "step": 3185
    },
    {
      "entropy": 6.329449844360352,
      "epoch": 0.28170257859413633,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995777302468628,
      "loss": 6.1553,
      "mean_token_accuracy": 0.1528245285153389,
      "num_tokens": 5949631.0,
      "step": 3190
    },
    {
      "entropy": 6.2650110721588135,
      "epoch": 0.28214411868597666,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995757995975789,
      "loss": 6.0648,
      "mean_token_accuracy": 0.16371494084596633,
      "num_tokens": 5957989.0,
      "step": 3195
    },
    {
      "entropy": 6.35319299697876,
      "epoch": 0.28258565877781705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995738645489828,
      "loss": 6.1873,
      "mean_token_accuracy": 0.14235186725854873,
      "num_tokens": 5968230.0,
      "step": 3200
    },
    {
      "entropy": 6.382534170150757,
      "epoch": 0.2830271988696574,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995719251011124,
      "loss": 6.2534,
      "mean_token_accuracy": 0.14914512187242507,
      "num_tokens": 5979379.0,
      "step": 3205
    },
    {
      "entropy": 6.341637182235718,
      "epoch": 0.2834687389614977,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995699812540058,
      "loss": 6.1775,
      "mean_token_accuracy": 0.15582906752824782,
      "num_tokens": 5988793.0,
      "step": 3210
    },
    {
      "entropy": 6.395340347290039,
      "epoch": 0.28391027905333804,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499568033007701,
      "loss": 6.2078,
      "mean_token_accuracy": 0.14774657115340234,
      "num_tokens": 5998439.0,
      "step": 3215
    },
    {
      "entropy": 6.244794750213623,
      "epoch": 0.2843518191451784,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995660803622361,
      "loss": 6.099,
      "mean_token_accuracy": 0.14922754764556884,
      "num_tokens": 6007113.0,
      "step": 3220
    },
    {
      "entropy": 6.318194818496704,
      "epoch": 0.2847933592370187,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995641233176494,
      "loss": 6.1506,
      "mean_token_accuracy": 0.14881237596273422,
      "num_tokens": 6015875.0,
      "step": 3225
    },
    {
      "entropy": 6.4028942584991455,
      "epoch": 0.28523489932885904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995621618739792,
      "loss": 6.2006,
      "mean_token_accuracy": 0.1527237117290497,
      "num_tokens": 6024717.0,
      "step": 3230
    },
    {
      "entropy": 6.235266828536988,
      "epoch": 0.2856764394206994,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499560196031264,
      "loss": 6.0941,
      "mean_token_accuracy": 0.15124002397060393,
      "num_tokens": 6032951.0,
      "step": 3235
    },
    {
      "entropy": 6.31634202003479,
      "epoch": 0.28611797951253976,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995582257895423,
      "loss": 6.1457,
      "mean_token_accuracy": 0.15818460434675216,
      "num_tokens": 6041969.0,
      "step": 3240
    },
    {
      "entropy": 6.215688705444336,
      "epoch": 0.2865595196043801,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995562511488528,
      "loss": 6.0334,
      "mean_token_accuracy": 0.16020716428756715,
      "num_tokens": 6050403.0,
      "step": 3245
    },
    {
      "entropy": 6.292816066741944,
      "epoch": 0.2870010596962204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995542721092337,
      "loss": 6.1135,
      "mean_token_accuracy": 0.1590289741754532,
      "num_tokens": 6058769.0,
      "step": 3250
    },
    {
      "entropy": 6.339272832870483,
      "epoch": 0.28744259978806075,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995522886707244,
      "loss": 6.1368,
      "mean_token_accuracy": 0.14725267589092256,
      "num_tokens": 6068342.0,
      "step": 3255
    },
    {
      "entropy": 6.207228803634644,
      "epoch": 0.2878841398799011,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004995503008333634,
      "loss": 6.0711,
      "mean_token_accuracy": 0.1586859941482544,
      "num_tokens": 6077331.0,
      "step": 3260
    },
    {
      "entropy": 6.35828013420105,
      "epoch": 0.2883256799717414,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995483085971897,
      "loss": 6.2419,
      "mean_token_accuracy": 0.14154957607388496,
      "num_tokens": 6087314.0,
      "step": 3265
    },
    {
      "entropy": 6.407521057128906,
      "epoch": 0.2887672200635818,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995463119622424,
      "loss": 6.1384,
      "mean_token_accuracy": 0.15381087362766266,
      "num_tokens": 6097046.0,
      "step": 3270
    },
    {
      "entropy": 6.2068273544311525,
      "epoch": 0.28920876015542213,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995443109285604,
      "loss": 6.0058,
      "mean_token_accuracy": 0.15988703668117524,
      "num_tokens": 6105999.0,
      "step": 3275
    },
    {
      "entropy": 6.266110754013061,
      "epoch": 0.28965030024726246,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995423054961832,
      "loss": 6.0711,
      "mean_token_accuracy": 0.15191173553466797,
      "num_tokens": 6113412.0,
      "step": 3280
    },
    {
      "entropy": 6.15813798904419,
      "epoch": 0.2900918403391028,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049954029566515,
      "loss": 6.0891,
      "mean_token_accuracy": 0.15964950025081634,
      "num_tokens": 6122515.0,
      "step": 3285
    },
    {
      "entropy": 6.3668163299560545,
      "epoch": 0.2905333804309431,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995382814355,
      "loss": 6.1651,
      "mean_token_accuracy": 0.15322745591402054,
      "num_tokens": 6131735.0,
      "step": 3290
    },
    {
      "entropy": 6.379836893081665,
      "epoch": 0.29097492052278345,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995362628072728,
      "loss": 6.1513,
      "mean_token_accuracy": 0.14620450288057327,
      "num_tokens": 6140556.0,
      "step": 3295
    },
    {
      "entropy": 6.208018493652344,
      "epoch": 0.2914164606146238,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995342397805078,
      "loss": 6.0491,
      "mean_token_accuracy": 0.16249649673700334,
      "num_tokens": 6149643.0,
      "step": 3300
    },
    {
      "entropy": 6.234081649780274,
      "epoch": 0.29185800070646417,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995322123552448,
      "loss": 6.1072,
      "mean_token_accuracy": 0.15427073538303376,
      "num_tokens": 6158941.0,
      "step": 3305
    },
    {
      "entropy": 6.347509908676147,
      "epoch": 0.2922995407983045,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995301805315235,
      "loss": 6.1508,
      "mean_token_accuracy": 0.150118388235569,
      "num_tokens": 6167975.0,
      "step": 3310
    },
    {
      "entropy": 6.244354009628296,
      "epoch": 0.29274108089014483,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995281443093837,
      "loss": 6.0,
      "mean_token_accuracy": 0.1684304043650627,
      "num_tokens": 6176758.0,
      "step": 3315
    },
    {
      "entropy": 6.2396119117736815,
      "epoch": 0.29318262098198516,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995261036888652,
      "loss": 6.1582,
      "mean_token_accuracy": 0.15510485470294952,
      "num_tokens": 6186348.0,
      "step": 3320
    },
    {
      "entropy": 6.349655723571777,
      "epoch": 0.2936241610738255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995240586700081,
      "loss": 6.1028,
      "mean_token_accuracy": 0.15705639719963074,
      "num_tokens": 6194646.0,
      "step": 3325
    },
    {
      "entropy": 6.28270206451416,
      "epoch": 0.2940657011656658,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995220092528522,
      "loss": 6.1965,
      "mean_token_accuracy": 0.14585646539926528,
      "num_tokens": 6204114.0,
      "step": 3330
    },
    {
      "entropy": 6.322914981842041,
      "epoch": 0.29450724125750616,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499519955437438,
      "loss": 6.1009,
      "mean_token_accuracy": 0.16212768703699112,
      "num_tokens": 6213473.0,
      "step": 3335
    },
    {
      "entropy": 6.2499192237854,
      "epoch": 0.29494878134934654,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995178972238054,
      "loss": 6.0633,
      "mean_token_accuracy": 0.16052547693252564,
      "num_tokens": 6221966.0,
      "step": 3340
    },
    {
      "entropy": 6.194470739364624,
      "epoch": 0.2953903214411869,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499515834611995,
      "loss": 6.0099,
      "mean_token_accuracy": 0.15819587111473082,
      "num_tokens": 6230401.0,
      "step": 3345
    },
    {
      "entropy": 6.300949764251709,
      "epoch": 0.2958318615330272,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995137676020472,
      "loss": 6.1786,
      "mean_token_accuracy": 0.1449178785085678,
      "num_tokens": 6240926.0,
      "step": 3350
    },
    {
      "entropy": 6.360338401794434,
      "epoch": 0.29627340162486754,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995116961940023,
      "loss": 6.1072,
      "mean_token_accuracy": 0.15931327491998673,
      "num_tokens": 6249904.0,
      "step": 3355
    },
    {
      "entropy": 6.252490758895874,
      "epoch": 0.29671494171670787,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995096203879009,
      "loss": 6.1643,
      "mean_token_accuracy": 0.15404981821775438,
      "num_tokens": 6258471.0,
      "step": 3360
    },
    {
      "entropy": 6.273584985733033,
      "epoch": 0.2971564818085482,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995075401837837,
      "loss": 6.0094,
      "mean_token_accuracy": 0.15978663712739943,
      "num_tokens": 6267036.0,
      "step": 3365
    },
    {
      "entropy": 6.260393381118774,
      "epoch": 0.29759802190038853,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995054555816915,
      "loss": 6.0878,
      "mean_token_accuracy": 0.16349994540214538,
      "num_tokens": 6276470.0,
      "step": 3370
    },
    {
      "entropy": 6.268303489685058,
      "epoch": 0.2980395619922289,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995033665816651,
      "loss": 6.0938,
      "mean_token_accuracy": 0.15316409915685653,
      "num_tokens": 6285466.0,
      "step": 3375
    },
    {
      "entropy": 6.22906641960144,
      "epoch": 0.29848110208406925,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995012731837454,
      "loss": 5.9686,
      "mean_token_accuracy": 0.16721353828907012,
      "num_tokens": 6294239.0,
      "step": 3380
    },
    {
      "entropy": 6.257667350769043,
      "epoch": 0.2989226421759096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994991753879736,
      "loss": 6.081,
      "mean_token_accuracy": 0.1589038774371147,
      "num_tokens": 6303170.0,
      "step": 3385
    },
    {
      "entropy": 6.261808824539185,
      "epoch": 0.2993641822677499,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994970731943904,
      "loss": 6.0758,
      "mean_token_accuracy": 0.15692856311798095,
      "num_tokens": 6312454.0,
      "step": 3390
    },
    {
      "entropy": 6.296799230575561,
      "epoch": 0.29980572235959024,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994949666030374,
      "loss": 6.1261,
      "mean_token_accuracy": 0.1539533942937851,
      "num_tokens": 6321947.0,
      "step": 3395
    },
    {
      "entropy": 6.218962526321411,
      "epoch": 0.3002472624514306,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994928556139557,
      "loss": 6.0338,
      "mean_token_accuracy": 0.16010112017393113,
      "num_tokens": 6330532.0,
      "step": 3400
    },
    {
      "entropy": 6.327949905395508,
      "epoch": 0.3006888025432709,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004994907402271865,
      "loss": 6.2111,
      "mean_token_accuracy": 0.15179989710450173,
      "num_tokens": 6340602.0,
      "step": 3405
    },
    {
      "entropy": 6.39948878288269,
      "epoch": 0.3011303426351113,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994886204427715,
      "loss": 6.1883,
      "mean_token_accuracy": 0.15149745345115662,
      "num_tokens": 6350522.0,
      "step": 3410
    },
    {
      "entropy": 6.317005920410156,
      "epoch": 0.3015718827269516,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994864962607519,
      "loss": 6.209,
      "mean_token_accuracy": 0.14751399978995322,
      "num_tokens": 6360796.0,
      "step": 3415
    },
    {
      "entropy": 6.284945154190064,
      "epoch": 0.30201342281879195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994843676811697,
      "loss": 6.0219,
      "mean_token_accuracy": 0.16533567160367965,
      "num_tokens": 6370624.0,
      "step": 3420
    },
    {
      "entropy": 6.244419860839844,
      "epoch": 0.3024549629106323,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994822347040664,
      "loss": 6.0812,
      "mean_token_accuracy": 0.1550205796957016,
      "num_tokens": 6379804.0,
      "step": 3425
    },
    {
      "entropy": 6.269517421722412,
      "epoch": 0.3028965030024726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994800973294837,
      "loss": 6.1065,
      "mean_token_accuracy": 0.15389672964811324,
      "num_tokens": 6388239.0,
      "step": 3430
    },
    {
      "entropy": 6.3750261783599855,
      "epoch": 0.30333804309431295,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994779555574636,
      "loss": 6.1595,
      "mean_token_accuracy": 0.15285229086875915,
      "num_tokens": 6398208.0,
      "step": 3435
    },
    {
      "entropy": 6.248434782028198,
      "epoch": 0.3037795831861533,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499475809388048,
      "loss": 6.0757,
      "mean_token_accuracy": 0.15916141718626023,
      "num_tokens": 6407701.0,
      "step": 3440
    },
    {
      "entropy": 6.227328872680664,
      "epoch": 0.30422112327799367,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499473658821279,
      "loss": 6.0083,
      "mean_token_accuracy": 0.15775404572486879,
      "num_tokens": 6416562.0,
      "step": 3445
    },
    {
      "entropy": 6.280228805541992,
      "epoch": 0.304662663369834,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994715038571986,
      "loss": 6.0821,
      "mean_token_accuracy": 0.16188960820436477,
      "num_tokens": 6425176.0,
      "step": 3450
    },
    {
      "entropy": 6.303675031661987,
      "epoch": 0.30510420346167433,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994693444958493,
      "loss": 6.0634,
      "mean_token_accuracy": 0.15637744665145875,
      "num_tokens": 6434911.0,
      "step": 3455
    },
    {
      "entropy": 6.213686847686768,
      "epoch": 0.30554574355351466,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499467180737273,
      "loss": 6.0409,
      "mean_token_accuracy": 0.16091670840978622,
      "num_tokens": 6444036.0,
      "step": 3460
    },
    {
      "entropy": 6.219314050674439,
      "epoch": 0.305987283645355,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994650125815124,
      "loss": 6.0956,
      "mean_token_accuracy": 0.15436640679836272,
      "num_tokens": 6453614.0,
      "step": 3465
    },
    {
      "entropy": 6.3259156227111815,
      "epoch": 0.3064288237371953,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994628400286097,
      "loss": 6.1891,
      "mean_token_accuracy": 0.15165584832429885,
      "num_tokens": 6463428.0,
      "step": 3470
    },
    {
      "entropy": 6.342681455612182,
      "epoch": 0.30687036382903565,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994606630786078,
      "loss": 6.0712,
      "mean_token_accuracy": 0.15518646240234374,
      "num_tokens": 6471985.0,
      "step": 3475
    },
    {
      "entropy": 6.2462522983551025,
      "epoch": 0.30731190392087604,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994584817315492,
      "loss": 6.1994,
      "mean_token_accuracy": 0.14526761025190355,
      "num_tokens": 6482590.0,
      "step": 3480
    },
    {
      "entropy": 6.304407978057862,
      "epoch": 0.30775344401271637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994562959874765,
      "loss": 6.2602,
      "mean_token_accuracy": 0.15167892575263978,
      "num_tokens": 6491981.0,
      "step": 3485
    },
    {
      "entropy": 6.3691192150115965,
      "epoch": 0.3081949841045567,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994541058464326,
      "loss": 6.1448,
      "mean_token_accuracy": 0.1520304873585701,
      "num_tokens": 6501256.0,
      "step": 3490
    },
    {
      "entropy": 6.31157956123352,
      "epoch": 0.30863652419639703,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994519113084605,
      "loss": 6.1322,
      "mean_token_accuracy": 0.15239002108573912,
      "num_tokens": 6509918.0,
      "step": 3495
    },
    {
      "entropy": 6.228223609924316,
      "epoch": 0.30907806428823736,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994497123736029,
      "loss": 6.0967,
      "mean_token_accuracy": 0.1539615049958229,
      "num_tokens": 6519725.0,
      "step": 3500
    },
    {
      "entropy": 6.288787221908569,
      "epoch": 0.3095196043800777,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994475090419034,
      "loss": 6.0995,
      "mean_token_accuracy": 0.15122905522584915,
      "num_tokens": 6528451.0,
      "step": 3505
    },
    {
      "entropy": 6.289122009277344,
      "epoch": 0.309961144471918,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994453013134047,
      "loss": 6.1775,
      "mean_token_accuracy": 0.1516020268201828,
      "num_tokens": 6538002.0,
      "step": 3510
    },
    {
      "entropy": 6.316947364807129,
      "epoch": 0.3104026845637584,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994430891881502,
      "loss": 6.0764,
      "mean_token_accuracy": 0.15740339010953902,
      "num_tokens": 6546583.0,
      "step": 3515
    },
    {
      "entropy": 6.344536685943604,
      "epoch": 0.31084422465559874,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994408726661832,
      "loss": 6.0918,
      "mean_token_accuracy": 0.15401178002357482,
      "num_tokens": 6556638.0,
      "step": 3520
    },
    {
      "entropy": 6.230458736419678,
      "epoch": 0.3112857647474391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994386517475472,
      "loss": 6.1739,
      "mean_token_accuracy": 0.1546187847852707,
      "num_tokens": 6566053.0,
      "step": 3525
    },
    {
      "entropy": 6.281338834762574,
      "epoch": 0.3117273048392794,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994364264322856,
      "loss": 6.166,
      "mean_token_accuracy": 0.14962629228830338,
      "num_tokens": 6575818.0,
      "step": 3530
    },
    {
      "entropy": 6.345117664337158,
      "epoch": 0.31216884493111974,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994341967204421,
      "loss": 6.1366,
      "mean_token_accuracy": 0.15367560982704162,
      "num_tokens": 6584210.0,
      "step": 3535
    },
    {
      "entropy": 6.276531553268432,
      "epoch": 0.31261038502296007,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994319626120603,
      "loss": 6.0339,
      "mean_token_accuracy": 0.16160414218902588,
      "num_tokens": 6592476.0,
      "step": 3540
    },
    {
      "entropy": 6.226123380661011,
      "epoch": 0.3130519251148004,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994297241071841,
      "loss": 6.0027,
      "mean_token_accuracy": 0.1591665893793106,
      "num_tokens": 6601299.0,
      "step": 3545
    },
    {
      "entropy": 6.2407856464385985,
      "epoch": 0.3134934652066408,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499427481205857,
      "loss": 6.0303,
      "mean_token_accuracy": 0.16062235981225967,
      "num_tokens": 6610808.0,
      "step": 3550
    },
    {
      "entropy": 6.247107124328613,
      "epoch": 0.3139350052984811,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994252339081234,
      "loss": 6.02,
      "mean_token_accuracy": 0.1675056204199791,
      "num_tokens": 6619700.0,
      "step": 3555
    },
    {
      "entropy": 6.185653591156006,
      "epoch": 0.31437654539032145,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499422982214027,
      "loss": 6.1187,
      "mean_token_accuracy": 0.15116817504167557,
      "num_tokens": 6628264.0,
      "step": 3560
    },
    {
      "entropy": 6.289248275756836,
      "epoch": 0.3148180854821618,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994207261236121,
      "loss": 6.0859,
      "mean_token_accuracy": 0.156384439766407,
      "num_tokens": 6637603.0,
      "step": 3565
    },
    {
      "entropy": 6.237731218338013,
      "epoch": 0.3152596255740021,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994184656369227,
      "loss": 6.1783,
      "mean_token_accuracy": 0.15488426461815835,
      "num_tokens": 6647736.0,
      "step": 3570
    },
    {
      "entropy": 6.2964622497558596,
      "epoch": 0.31570116566584244,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994162007540033,
      "loss": 6.07,
      "mean_token_accuracy": 0.16558655798435212,
      "num_tokens": 6656490.0,
      "step": 3575
    },
    {
      "entropy": 6.285622644424438,
      "epoch": 0.3161427057576828,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994139314748981,
      "loss": 6.072,
      "mean_token_accuracy": 0.1698063611984253,
      "num_tokens": 6666608.0,
      "step": 3580
    },
    {
      "entropy": 6.262637805938721,
      "epoch": 0.31658424584952316,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994116577996517,
      "loss": 6.1181,
      "mean_token_accuracy": 0.15824046060442926,
      "num_tokens": 6676167.0,
      "step": 3585
    },
    {
      "entropy": 6.249921607971191,
      "epoch": 0.3170257859413635,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994093797283084,
      "loss": 6.0807,
      "mean_token_accuracy": 0.154364275932312,
      "num_tokens": 6685527.0,
      "step": 3590
    },
    {
      "entropy": 6.2948345184326175,
      "epoch": 0.3174673260332038,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994070972609132,
      "loss": 6.1098,
      "mean_token_accuracy": 0.15264881104230882,
      "num_tokens": 6694273.0,
      "step": 3595
    },
    {
      "entropy": 6.215621471405029,
      "epoch": 0.31790886612504415,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994048103975103,
      "loss": 6.048,
      "mean_token_accuracy": 0.16299316436052322,
      "num_tokens": 6703276.0,
      "step": 3600
    },
    {
      "entropy": 6.296894073486328,
      "epoch": 0.3183504062168845,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499402519138145,
      "loss": 6.1125,
      "mean_token_accuracy": 0.1538652792572975,
      "num_tokens": 6713198.0,
      "step": 3605
    },
    {
      "entropy": 6.275954961776733,
      "epoch": 0.3187919463087248,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994002234828619,
      "loss": 6.1491,
      "mean_token_accuracy": 0.14983329325914382,
      "num_tokens": 6722721.0,
      "step": 3610
    },
    {
      "entropy": 6.340907716751099,
      "epoch": 0.31923348640056515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499397923431706,
      "loss": 6.0051,
      "mean_token_accuracy": 0.16155414134263993,
      "num_tokens": 6731312.0,
      "step": 3615
    },
    {
      "entropy": 6.223904609680176,
      "epoch": 0.31967502649240553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993956189847226,
      "loss": 6.0843,
      "mean_token_accuracy": 0.15967404693365098,
      "num_tokens": 6740444.0,
      "step": 3620
    },
    {
      "entropy": 6.326261425018311,
      "epoch": 0.32011656658424587,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993933101419565,
      "loss": 6.1317,
      "mean_token_accuracy": 0.14693555012345313,
      "num_tokens": 6749770.0,
      "step": 3625
    },
    {
      "entropy": 6.276262807846069,
      "epoch": 0.3205581066760862,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993909969034531,
      "loss": 6.0732,
      "mean_token_accuracy": 0.15638579428195953,
      "num_tokens": 6758713.0,
      "step": 3630
    },
    {
      "entropy": 6.221458292007446,
      "epoch": 0.3209996467679265,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993886792692576,
      "loss": 5.9939,
      "mean_token_accuracy": 0.161094631254673,
      "num_tokens": 6768309.0,
      "step": 3635
    },
    {
      "entropy": 6.214659118652344,
      "epoch": 0.32144118685976686,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993863572394156,
      "loss": 6.0857,
      "mean_token_accuracy": 0.1580155685544014,
      "num_tokens": 6777428.0,
      "step": 3640
    },
    {
      "entropy": 6.327162265777588,
      "epoch": 0.3218827269516072,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993840308139724,
      "loss": 6.0625,
      "mean_token_accuracy": 0.15655053928494453,
      "num_tokens": 6787263.0,
      "step": 3645
    },
    {
      "entropy": 6.2021434783935545,
      "epoch": 0.3223242670434475,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993816999929738,
      "loss": 6.0361,
      "mean_token_accuracy": 0.15779577121138572,
      "num_tokens": 6797789.0,
      "step": 3650
    },
    {
      "entropy": 6.212742614746094,
      "epoch": 0.3227658071352879,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993793647764651,
      "loss": 6.0641,
      "mean_token_accuracy": 0.16072517931461333,
      "num_tokens": 6808703.0,
      "step": 3655
    },
    {
      "entropy": 6.262935543060303,
      "epoch": 0.32320734722712824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993770251644923,
      "loss": 6.0083,
      "mean_token_accuracy": 0.16098256260156632,
      "num_tokens": 6817611.0,
      "step": 3660
    },
    {
      "entropy": 6.200230598449707,
      "epoch": 0.32364888731896857,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993746811571013,
      "loss": 6.0241,
      "mean_token_accuracy": 0.161439149081707,
      "num_tokens": 6827316.0,
      "step": 3665
    },
    {
      "entropy": 6.2396519660949705,
      "epoch": 0.3240904274108089,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993723327543379,
      "loss": 6.054,
      "mean_token_accuracy": 0.16056997030973436,
      "num_tokens": 6836360.0,
      "step": 3670
    },
    {
      "entropy": 6.310636806488037,
      "epoch": 0.32453196750264923,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499369979956248,
      "loss": 6.0777,
      "mean_token_accuracy": 0.1472184494137764,
      "num_tokens": 6845243.0,
      "step": 3675
    },
    {
      "entropy": 6.26262788772583,
      "epoch": 0.32497350759448956,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993676227628779,
      "loss": 5.9876,
      "mean_token_accuracy": 0.1617436170578003,
      "num_tokens": 6853610.0,
      "step": 3680
    },
    {
      "entropy": 6.162088060379029,
      "epoch": 0.3254150476863299,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993652611742736,
      "loss": 6.0497,
      "mean_token_accuracy": 0.16987728625535964,
      "num_tokens": 6862823.0,
      "step": 3685
    },
    {
      "entropy": 6.234931087493896,
      "epoch": 0.3258565877781703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993628951904815,
      "loss": 6.1277,
      "mean_token_accuracy": 0.14977747499942778,
      "num_tokens": 6871858.0,
      "step": 3690
    },
    {
      "entropy": 6.2675896167755125,
      "epoch": 0.3262981278700106,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993605248115479,
      "loss": 6.085,
      "mean_token_accuracy": 0.16308356672525406,
      "num_tokens": 6881275.0,
      "step": 3695
    },
    {
      "entropy": 6.220019578933716,
      "epoch": 0.32673966796185094,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993581500375191,
      "loss": 6.0395,
      "mean_token_accuracy": 0.1569608435034752,
      "num_tokens": 6889695.0,
      "step": 3700
    },
    {
      "entropy": 6.2498650550842285,
      "epoch": 0.3271812080536913,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993557708684417,
      "loss": 6.0661,
      "mean_token_accuracy": 0.1529826432466507,
      "num_tokens": 6898735.0,
      "step": 3705
    },
    {
      "entropy": 6.166424942016602,
      "epoch": 0.3276227481455316,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993533873043625,
      "loss": 5.9981,
      "mean_token_accuracy": 0.16759352684020995,
      "num_tokens": 6907864.0,
      "step": 3710
    },
    {
      "entropy": 6.264701557159424,
      "epoch": 0.32806428823737194,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499350999345328,
      "loss": 6.0913,
      "mean_token_accuracy": 0.1573421210050583,
      "num_tokens": 6917677.0,
      "step": 3715
    },
    {
      "entropy": 6.259223127365113,
      "epoch": 0.32850582832921227,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499348606991385,
      "loss": 6.0851,
      "mean_token_accuracy": 0.1528132177889347,
      "num_tokens": 6927060.0,
      "step": 3720
    },
    {
      "entropy": 6.232635593414306,
      "epoch": 0.32894736842105265,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993462102425805,
      "loss": 6.0696,
      "mean_token_accuracy": 0.15622131675481796,
      "num_tokens": 6936259.0,
      "step": 3725
    },
    {
      "entropy": 6.352907133102417,
      "epoch": 0.329388908512893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993438090989612,
      "loss": 6.0884,
      "mean_token_accuracy": 0.1543697103857994,
      "num_tokens": 6945468.0,
      "step": 3730
    },
    {
      "entropy": 6.245824527740479,
      "epoch": 0.3298304486047333,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993414035605743,
      "loss": 6.0411,
      "mean_token_accuracy": 0.15524725764989852,
      "num_tokens": 6955577.0,
      "step": 3735
    },
    {
      "entropy": 6.262871217727661,
      "epoch": 0.33027198869657365,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993389936274669,
      "loss": 6.1449,
      "mean_token_accuracy": 0.15381260961294174,
      "num_tokens": 6965003.0,
      "step": 3740
    },
    {
      "entropy": 6.235921144485474,
      "epoch": 0.330713528788414,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993365792996862,
      "loss": 6.1082,
      "mean_token_accuracy": 0.15733923763036728,
      "num_tokens": 6975501.0,
      "step": 3745
    },
    {
      "entropy": 6.247527360916138,
      "epoch": 0.3311550688802543,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993341605772795,
      "loss": 6.0996,
      "mean_token_accuracy": 0.1563773959875107,
      "num_tokens": 6984601.0,
      "step": 3750
    },
    {
      "entropy": 6.228341197967529,
      "epoch": 0.33159660897209464,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993317374602941,
      "loss": 6.0413,
      "mean_token_accuracy": 0.15963054150342942,
      "num_tokens": 6993918.0,
      "step": 3755
    },
    {
      "entropy": 6.260748291015625,
      "epoch": 0.33203814906393503,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993293099487777,
      "loss": 6.1194,
      "mean_token_accuracy": 0.1574609696865082,
      "num_tokens": 7003937.0,
      "step": 3760
    },
    {
      "entropy": 6.255482339859009,
      "epoch": 0.33247968915577536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993268780427776,
      "loss": 6.0727,
      "mean_token_accuracy": 0.15733179301023484,
      "num_tokens": 7013346.0,
      "step": 3765
    },
    {
      "entropy": 6.202826976776123,
      "epoch": 0.3329212292476157,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993244417423416,
      "loss": 5.916,
      "mean_token_accuracy": 0.16502410620450975,
      "num_tokens": 7022275.0,
      "step": 3770
    },
    {
      "entropy": 6.13905611038208,
      "epoch": 0.333362769339456,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993220010475174,
      "loss": 6.0295,
      "mean_token_accuracy": 0.15234270095825195,
      "num_tokens": 7031101.0,
      "step": 3775
    },
    {
      "entropy": 6.27298412322998,
      "epoch": 0.33380430943129635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993195559583526,
      "loss": 6.0261,
      "mean_token_accuracy": 0.15508327782154083,
      "num_tokens": 7039898.0,
      "step": 3780
    },
    {
      "entropy": 6.202321863174438,
      "epoch": 0.3342458495231367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993171064748954,
      "loss": 6.0459,
      "mean_token_accuracy": 0.1603365957736969,
      "num_tokens": 7048356.0,
      "step": 3785
    },
    {
      "entropy": 6.235737991333008,
      "epoch": 0.334687389614977,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993146525971937,
      "loss": 6.1182,
      "mean_token_accuracy": 0.1601865068078041,
      "num_tokens": 7058030.0,
      "step": 3790
    },
    {
      "entropy": 6.207295894622803,
      "epoch": 0.3351289297068174,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993121943252955,
      "loss": 6.0448,
      "mean_token_accuracy": 0.15825917571783066,
      "num_tokens": 7067864.0,
      "step": 3795
    },
    {
      "entropy": 6.2124467372894285,
      "epoch": 0.33557046979865773,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993097316592489,
      "loss": 5.9416,
      "mean_token_accuracy": 0.16415700912475586,
      "num_tokens": 7076408.0,
      "step": 3800
    },
    {
      "entropy": 6.1406035900115965,
      "epoch": 0.33601200989049806,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993072645991023,
      "loss": 5.9999,
      "mean_token_accuracy": 0.1686277985572815,
      "num_tokens": 7084608.0,
      "step": 3805
    },
    {
      "entropy": 6.227652168273925,
      "epoch": 0.3364535499823384,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499304793144904,
      "loss": 6.0712,
      "mean_token_accuracy": 0.1554234266281128,
      "num_tokens": 7093332.0,
      "step": 3810
    },
    {
      "entropy": 6.224140739440918,
      "epoch": 0.3368950900741787,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993023172967022,
      "loss": 6.0269,
      "mean_token_accuracy": 0.16415185630321502,
      "num_tokens": 7102223.0,
      "step": 3815
    },
    {
      "entropy": 6.212011909484863,
      "epoch": 0.33733663016601906,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992998370545458,
      "loss": 6.073,
      "mean_token_accuracy": 0.15606720745563507,
      "num_tokens": 7112083.0,
      "step": 3820
    },
    {
      "entropy": 6.2539793968200685,
      "epoch": 0.3377781702578594,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992973524184831,
      "loss": 5.9761,
      "mean_token_accuracy": 0.15876844376325608,
      "num_tokens": 7121383.0,
      "step": 3825
    },
    {
      "entropy": 6.223515558242798,
      "epoch": 0.3382197103496998,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992948633885627,
      "loss": 6.1649,
      "mean_token_accuracy": 0.1447250097990036,
      "num_tokens": 7130288.0,
      "step": 3830
    },
    {
      "entropy": 6.2508656024932865,
      "epoch": 0.3386612504415401,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992923699648335,
      "loss": 6.0337,
      "mean_token_accuracy": 0.1640491470694542,
      "num_tokens": 7139135.0,
      "step": 3835
    },
    {
      "entropy": 6.262220048904419,
      "epoch": 0.33910279053338044,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992898721473445,
      "loss": 6.0715,
      "mean_token_accuracy": 0.15599901527166365,
      "num_tokens": 7148184.0,
      "step": 3840
    },
    {
      "entropy": 6.127392101287842,
      "epoch": 0.33954433062522077,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992873699361444,
      "loss": 6.0425,
      "mean_token_accuracy": 0.15902995243668555,
      "num_tokens": 7157258.0,
      "step": 3845
    },
    {
      "entropy": 6.235959815979004,
      "epoch": 0.3399858707170611,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992848633312822,
      "loss": 6.0776,
      "mean_token_accuracy": 0.1621302619576454,
      "num_tokens": 7167375.0,
      "step": 3850
    },
    {
      "entropy": 6.30757737159729,
      "epoch": 0.34042741080890143,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992823523328071,
      "loss": 5.9866,
      "mean_token_accuracy": 0.1621929094195366,
      "num_tokens": 7176400.0,
      "step": 3855
    },
    {
      "entropy": 6.13206639289856,
      "epoch": 0.34086895090074176,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992798369407684,
      "loss": 6.0313,
      "mean_token_accuracy": 0.15993043184280395,
      "num_tokens": 7187125.0,
      "step": 3860
    },
    {
      "entropy": 6.272826147079468,
      "epoch": 0.34131049099258215,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992773171552152,
      "loss": 6.0365,
      "mean_token_accuracy": 0.16591625660657883,
      "num_tokens": 7196876.0,
      "step": 3865
    },
    {
      "entropy": 6.263292264938355,
      "epoch": 0.3417520310844225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992747929761968,
      "loss": 6.066,
      "mean_token_accuracy": 0.1596837028861046,
      "num_tokens": 7205595.0,
      "step": 3870
    },
    {
      "entropy": 6.227126216888427,
      "epoch": 0.3421935711762628,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992722644037628,
      "loss": 6.0107,
      "mean_token_accuracy": 0.15742239356040955,
      "num_tokens": 7215741.0,
      "step": 3875
    },
    {
      "entropy": 6.259053897857666,
      "epoch": 0.34263511126810314,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992697314379628,
      "loss": 6.0913,
      "mean_token_accuracy": 0.15837327837944032,
      "num_tokens": 7225300.0,
      "step": 3880
    },
    {
      "entropy": 6.19787712097168,
      "epoch": 0.3430766513599435,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992671940788462,
      "loss": 6.0593,
      "mean_token_accuracy": 0.15788772851228713,
      "num_tokens": 7234474.0,
      "step": 3885
    },
    {
      "entropy": 6.092776584625244,
      "epoch": 0.3435181914517838,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992646523264628,
      "loss": 5.9614,
      "mean_token_accuracy": 0.1593474417924881,
      "num_tokens": 7244441.0,
      "step": 3890
    },
    {
      "entropy": 6.190795946121216,
      "epoch": 0.34395973154362414,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992621061808625,
      "loss": 6.0144,
      "mean_token_accuracy": 0.16093249768018722,
      "num_tokens": 7253260.0,
      "step": 3895
    },
    {
      "entropy": 6.322411632537841,
      "epoch": 0.3444012716354645,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499259555642095,
      "loss": 6.2059,
      "mean_token_accuracy": 0.15132326781749725,
      "num_tokens": 7262565.0,
      "step": 3900
    },
    {
      "entropy": 6.1810626029968265,
      "epoch": 0.34484281172730485,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992570007102104,
      "loss": 6.0234,
      "mean_token_accuracy": 0.15467487573623656,
      "num_tokens": 7270789.0,
      "step": 3905
    },
    {
      "entropy": 6.2457053661346436,
      "epoch": 0.3452843518191452,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992544413852587,
      "loss": 6.1059,
      "mean_token_accuracy": 0.15839382708072663,
      "num_tokens": 7279366.0,
      "step": 3910
    },
    {
      "entropy": 6.228488111495972,
      "epoch": 0.3457258919109855,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049925187766729,
      "loss": 6.0294,
      "mean_token_accuracy": 0.15858375430107116,
      "num_tokens": 7288916.0,
      "step": 3915
    },
    {
      "entropy": 6.239999008178711,
      "epoch": 0.34616743200282585,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992493095563545,
      "loss": 6.1154,
      "mean_token_accuracy": 0.15214944928884505,
      "num_tokens": 7299732.0,
      "step": 3920
    },
    {
      "entropy": 6.32588939666748,
      "epoch": 0.3466089720946662,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992467370525026,
      "loss": 6.165,
      "mean_token_accuracy": 0.14919357597827912,
      "num_tokens": 7309601.0,
      "step": 3925
    },
    {
      "entropy": 6.196690225601197,
      "epoch": 0.3470505121865065,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992441601557848,
      "loss": 5.9955,
      "mean_token_accuracy": 0.16536309123039244,
      "num_tokens": 7318949.0,
      "step": 3930
    },
    {
      "entropy": 6.181383752822876,
      "epoch": 0.3474920522783469,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992415788662514,
      "loss": 5.9967,
      "mean_token_accuracy": 0.16176612675189972,
      "num_tokens": 7327885.0,
      "step": 3935
    },
    {
      "entropy": 6.27258563041687,
      "epoch": 0.34793359237018723,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992389931839529,
      "loss": 6.1387,
      "mean_token_accuracy": 0.14940299838781357,
      "num_tokens": 7337194.0,
      "step": 3940
    },
    {
      "entropy": 6.264700651168823,
      "epoch": 0.34837513246202756,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992364031089401,
      "loss": 5.9928,
      "mean_token_accuracy": 0.15456721186637878,
      "num_tokens": 7346193.0,
      "step": 3945
    },
    {
      "entropy": 6.093889999389648,
      "epoch": 0.3488166725538679,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992338086412636,
      "loss": 6.0922,
      "mean_token_accuracy": 0.15651808083057403,
      "num_tokens": 7355017.0,
      "step": 3950
    },
    {
      "entropy": 6.302392911911011,
      "epoch": 0.3492582126457082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992312097809744,
      "loss": 6.0613,
      "mean_token_accuracy": 0.16095637679100036,
      "num_tokens": 7364736.0,
      "step": 3955
    },
    {
      "entropy": 6.133909893035889,
      "epoch": 0.34969975273754855,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992286065281234,
      "loss": 5.9467,
      "mean_token_accuracy": 0.16424252539873124,
      "num_tokens": 7374141.0,
      "step": 3960
    },
    {
      "entropy": 6.1777966022491455,
      "epoch": 0.3501412928293889,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992259988827614,
      "loss": 5.921,
      "mean_token_accuracy": 0.1659179836511612,
      "num_tokens": 7383124.0,
      "step": 3965
    },
    {
      "entropy": 6.174456310272217,
      "epoch": 0.35058283292122927,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992233868449397,
      "loss": 5.9251,
      "mean_token_accuracy": 0.16522225886583328,
      "num_tokens": 7392736.0,
      "step": 3970
    },
    {
      "entropy": 6.207459211349487,
      "epoch": 0.3510243730130696,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992207704147093,
      "loss": 6.0076,
      "mean_token_accuracy": 0.16019548624753951,
      "num_tokens": 7401354.0,
      "step": 3975
    },
    {
      "entropy": 6.130263948440552,
      "epoch": 0.35146591310490993,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992181495921216,
      "loss": 5.996,
      "mean_token_accuracy": 0.1623647540807724,
      "num_tokens": 7410306.0,
      "step": 3980
    },
    {
      "entropy": 6.275353193283081,
      "epoch": 0.35190745319675026,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992155243772277,
      "loss": 6.0688,
      "mean_token_accuracy": 0.15552889704704284,
      "num_tokens": 7420422.0,
      "step": 3985
    },
    {
      "entropy": 6.313285398483276,
      "epoch": 0.3523489932885906,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992128947700795,
      "loss": 6.1209,
      "mean_token_accuracy": 0.1604642853140831,
      "num_tokens": 7430161.0,
      "step": 3990
    },
    {
      "entropy": 6.132496166229248,
      "epoch": 0.3527905333804309,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499210260770728,
      "loss": 6.0426,
      "mean_token_accuracy": 0.16399580985307693,
      "num_tokens": 7439164.0,
      "step": 3995
    },
    {
      "entropy": 6.244388103485107,
      "epoch": 0.35323207347227126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499207622379225,
      "loss": 6.1026,
      "mean_token_accuracy": 0.15584832578897476,
      "num_tokens": 7448446.0,
      "step": 4000
    },
    {
      "entropy": 6.1597412109375,
      "epoch": 0.35367361356411164,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992049795956222,
      "loss": 5.9986,
      "mean_token_accuracy": 0.16560217887163162,
      "num_tokens": 7457564.0,
      "step": 4005
    },
    {
      "entropy": 6.190818500518799,
      "epoch": 0.354115153655952,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992023324199715,
      "loss": 5.9764,
      "mean_token_accuracy": 0.1704181656241417,
      "num_tokens": 7467198.0,
      "step": 4010
    },
    {
      "entropy": 6.166417837142944,
      "epoch": 0.3545566937477923,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991996808523245,
      "loss": 6.0217,
      "mean_token_accuracy": 0.1554143711924553,
      "num_tokens": 7477052.0,
      "step": 4015
    },
    {
      "entropy": 6.247974443435669,
      "epoch": 0.35499823383963264,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991970248927332,
      "loss": 5.9991,
      "mean_token_accuracy": 0.15834062546491623,
      "num_tokens": 7486999.0,
      "step": 4020
    },
    {
      "entropy": 6.197288513183594,
      "epoch": 0.35543977393147297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991943645412498,
      "loss": 6.0052,
      "mean_token_accuracy": 0.15766916126012803,
      "num_tokens": 7495791.0,
      "step": 4025
    },
    {
      "entropy": 6.1609608173370365,
      "epoch": 0.3558813140233133,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991916997979263,
      "loss": 5.9389,
      "mean_token_accuracy": 0.16242348700761794,
      "num_tokens": 7505002.0,
      "step": 4030
    },
    {
      "entropy": 6.192747592926025,
      "epoch": 0.35632285411515363,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991890306628149,
      "loss": 5.9854,
      "mean_token_accuracy": 0.15994453579187393,
      "num_tokens": 7514001.0,
      "step": 4035
    },
    {
      "entropy": 6.133841800689697,
      "epoch": 0.356764394206994,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991863571359678,
      "loss": 5.9624,
      "mean_token_accuracy": 0.1689068853855133,
      "num_tokens": 7522854.0,
      "step": 4040
    },
    {
      "entropy": 6.153005123138428,
      "epoch": 0.35720593429883435,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991836792174376,
      "loss": 6.0649,
      "mean_token_accuracy": 0.15469688326120376,
      "num_tokens": 7532220.0,
      "step": 4045
    },
    {
      "entropy": 6.2448680877685545,
      "epoch": 0.3576474743906747,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991809969072765,
      "loss": 5.9876,
      "mean_token_accuracy": 0.16091796904802322,
      "num_tokens": 7541013.0,
      "step": 4050
    },
    {
      "entropy": 6.135132312774658,
      "epoch": 0.358089014482515,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991783102055371,
      "loss": 5.9553,
      "mean_token_accuracy": 0.16446686983108522,
      "num_tokens": 7550746.0,
      "step": 4055
    },
    {
      "entropy": 6.128793001174927,
      "epoch": 0.35853055457435534,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991756191122723,
      "loss": 5.9993,
      "mean_token_accuracy": 0.15976244807243348,
      "num_tokens": 7560009.0,
      "step": 4060
    },
    {
      "entropy": 6.115909385681152,
      "epoch": 0.3589720946661957,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991729236275346,
      "loss": 5.959,
      "mean_token_accuracy": 0.1665020152926445,
      "num_tokens": 7568190.0,
      "step": 4065
    },
    {
      "entropy": 6.270355844497681,
      "epoch": 0.359413634758036,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991702237513768,
      "loss": 6.0048,
      "mean_token_accuracy": 0.16158075258135796,
      "num_tokens": 7577535.0,
      "step": 4070
    },
    {
      "entropy": 6.24889907836914,
      "epoch": 0.3598551748498764,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991675194838517,
      "loss": 6.0622,
      "mean_token_accuracy": 0.15503631234169007,
      "num_tokens": 7588268.0,
      "step": 4075
    },
    {
      "entropy": 6.159323072433471,
      "epoch": 0.3602967149417167,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991648108250125,
      "loss": 5.9068,
      "mean_token_accuracy": 0.16366464048624038,
      "num_tokens": 7596966.0,
      "step": 4080
    },
    {
      "entropy": 6.191630887985229,
      "epoch": 0.36073825503355705,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499162097774912,
      "loss": 6.0638,
      "mean_token_accuracy": 0.16405229419469833,
      "num_tokens": 7606110.0,
      "step": 4085
    },
    {
      "entropy": 6.204769897460937,
      "epoch": 0.3611797951253974,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991593803336037,
      "loss": 5.9683,
      "mean_token_accuracy": 0.1623123914003372,
      "num_tokens": 7614865.0,
      "step": 4090
    },
    {
      "entropy": 6.260302019119263,
      "epoch": 0.3616213352172377,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991566585011405,
      "loss": 6.065,
      "mean_token_accuracy": 0.15668002367019654,
      "num_tokens": 7624468.0,
      "step": 4095
    },
    {
      "entropy": 6.122881031036377,
      "epoch": 0.36206287530907805,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004991539322775758,
      "loss": 6.0627,
      "mean_token_accuracy": 0.16159093379974365,
      "num_tokens": 7634173.0,
      "step": 4100
    },
    {
      "entropy": 6.228496980667114,
      "epoch": 0.3625044154009184,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991512016629632,
      "loss": 5.9559,
      "mean_token_accuracy": 0.1651410400867462,
      "num_tokens": 7643171.0,
      "step": 4105
    },
    {
      "entropy": 6.2540699481964115,
      "epoch": 0.36294595549275877,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991484666573558,
      "loss": 6.0,
      "mean_token_accuracy": 0.15876996368169785,
      "num_tokens": 7651582.0,
      "step": 4110
    },
    {
      "entropy": 6.151803731918335,
      "epoch": 0.3633874955845991,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004991457272608077,
      "loss": 6.0921,
      "mean_token_accuracy": 0.15867114216089248,
      "num_tokens": 7661697.0,
      "step": 4115
    },
    {
      "entropy": 6.168854808807373,
      "epoch": 0.3638290356764394,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991429834733721,
      "loss": 5.9076,
      "mean_token_accuracy": 0.1628608524799347,
      "num_tokens": 7670619.0,
      "step": 4120
    },
    {
      "entropy": 6.13516321182251,
      "epoch": 0.36427057576827976,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499140235295103,
      "loss": 5.9742,
      "mean_token_accuracy": 0.1674422413110733,
      "num_tokens": 7678937.0,
      "step": 4125
    },
    {
      "entropy": 6.178045797348022,
      "epoch": 0.3647121158601201,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991374827260542,
      "loss": 6.009,
      "mean_token_accuracy": 0.160325326025486,
      "num_tokens": 7688925.0,
      "step": 4130
    },
    {
      "entropy": 6.165687131881714,
      "epoch": 0.3651536559519604,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991347257662795,
      "loss": 5.9925,
      "mean_token_accuracy": 0.16649454236030578,
      "num_tokens": 7699397.0,
      "step": 4135
    },
    {
      "entropy": 6.169813394546509,
      "epoch": 0.36559519604380075,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499131964415833,
      "loss": 6.086,
      "mean_token_accuracy": 0.15795015543699265,
      "num_tokens": 7709205.0,
      "step": 4140
    },
    {
      "entropy": 6.220038318634034,
      "epoch": 0.36603673613564114,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991291986747689,
      "loss": 6.0146,
      "mean_token_accuracy": 0.1661527544260025,
      "num_tokens": 7717622.0,
      "step": 4145
    },
    {
      "entropy": 6.187426042556763,
      "epoch": 0.36647827622748147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991264285431412,
      "loss": 6.0405,
      "mean_token_accuracy": 0.16048768162727356,
      "num_tokens": 7727795.0,
      "step": 4150
    },
    {
      "entropy": 6.184950828552246,
      "epoch": 0.3669198163193218,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991236540210041,
      "loss": 5.9323,
      "mean_token_accuracy": 0.164162714779377,
      "num_tokens": 7736539.0,
      "step": 4155
    },
    {
      "entropy": 6.1612285614013675,
      "epoch": 0.36736135641116213,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991208751084122,
      "loss": 5.9621,
      "mean_token_accuracy": 0.16195394843816757,
      "num_tokens": 7745660.0,
      "step": 4160
    },
    {
      "entropy": 6.126555776596069,
      "epoch": 0.36780289650300246,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004991180918054199,
      "loss": 6.0392,
      "mean_token_accuracy": 0.15680405348539353,
      "num_tokens": 7755257.0,
      "step": 4165
    },
    {
      "entropy": 6.23470664024353,
      "epoch": 0.3682444365948428,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991153041120815,
      "loss": 5.925,
      "mean_token_accuracy": 0.16362495571374894,
      "num_tokens": 7764140.0,
      "step": 4170
    },
    {
      "entropy": 6.117575931549072,
      "epoch": 0.3686859766866831,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991125120284519,
      "loss": 5.9156,
      "mean_token_accuracy": 0.16395330131053926,
      "num_tokens": 7772404.0,
      "step": 4175
    },
    {
      "entropy": 6.117729043960571,
      "epoch": 0.3691275167785235,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991097155545856,
      "loss": 5.9434,
      "mean_token_accuracy": 0.16438591182231904,
      "num_tokens": 7781321.0,
      "step": 4180
    },
    {
      "entropy": 6.221078300476075,
      "epoch": 0.36956905687036384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991069146905374,
      "loss": 6.0666,
      "mean_token_accuracy": 0.16164010018110275,
      "num_tokens": 7789825.0,
      "step": 4185
    },
    {
      "entropy": 6.228800964355469,
      "epoch": 0.3700105969622042,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991041094363621,
      "loss": 6.0397,
      "mean_token_accuracy": 0.16061489135026932,
      "num_tokens": 7798536.0,
      "step": 4190
    },
    {
      "entropy": 6.155390214920044,
      "epoch": 0.3704521370540445,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991012997921149,
      "loss": 6.0116,
      "mean_token_accuracy": 0.15706663131713866,
      "num_tokens": 7808213.0,
      "step": 4195
    },
    {
      "entropy": 6.2576686382293705,
      "epoch": 0.37089367714588484,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990984857578506,
      "loss": 6.0799,
      "mean_token_accuracy": 0.15308379232883454,
      "num_tokens": 7817895.0,
      "step": 4200
    },
    {
      "entropy": 6.221368598937988,
      "epoch": 0.37133521723772517,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990956673336245,
      "loss": 6.0392,
      "mean_token_accuracy": 0.15452028661966324,
      "num_tokens": 7827019.0,
      "step": 4205
    },
    {
      "entropy": 6.151476812362671,
      "epoch": 0.3717767573295655,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990928445194917,
      "loss": 6.0149,
      "mean_token_accuracy": 0.16029624193906783,
      "num_tokens": 7835938.0,
      "step": 4210
    },
    {
      "entropy": 6.1919189453125,
      "epoch": 0.3722182974214059,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990900173155074,
      "loss": 6.0267,
      "mean_token_accuracy": 0.16564586609601975,
      "num_tokens": 7845776.0,
      "step": 4215
    },
    {
      "entropy": 6.0918032169342045,
      "epoch": 0.3726598375132462,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990871857217273,
      "loss": 5.9565,
      "mean_token_accuracy": 0.1695231392979622,
      "num_tokens": 7855573.0,
      "step": 4220
    },
    {
      "entropy": 6.173537826538086,
      "epoch": 0.37310137760508655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990843497382066,
      "loss": 6.0679,
      "mean_token_accuracy": 0.16159802228212355,
      "num_tokens": 7866611.0,
      "step": 4225
    },
    {
      "entropy": 6.1972291469573975,
      "epoch": 0.3735429176969269,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990815093650009,
      "loss": 5.9074,
      "mean_token_accuracy": 0.17359337508678435,
      "num_tokens": 7875638.0,
      "step": 4230
    },
    {
      "entropy": 6.115181493759155,
      "epoch": 0.3739844577887672,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990786646021659,
      "loss": 6.0433,
      "mean_token_accuracy": 0.15696808993816375,
      "num_tokens": 7884800.0,
      "step": 4235
    },
    {
      "entropy": 6.139626598358154,
      "epoch": 0.37442599788060754,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990758154497573,
      "loss": 5.9302,
      "mean_token_accuracy": 0.17153851687908173,
      "num_tokens": 7894676.0,
      "step": 4240
    },
    {
      "entropy": 6.202898693084717,
      "epoch": 0.3748675379724479,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990729619078309,
      "loss": 6.0135,
      "mean_token_accuracy": 0.1623058333992958,
      "num_tokens": 7904130.0,
      "step": 4245
    },
    {
      "entropy": 6.16353211402893,
      "epoch": 0.37530907806428826,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990701039764427,
      "loss": 5.9502,
      "mean_token_accuracy": 0.16358196958899499,
      "num_tokens": 7913334.0,
      "step": 4250
    },
    {
      "entropy": 6.283150625228882,
      "epoch": 0.3757506181561286,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990672416556485,
      "loss": 6.0754,
      "mean_token_accuracy": 0.16275528818368912,
      "num_tokens": 7922718.0,
      "step": 4255
    },
    {
      "entropy": 6.174107217788697,
      "epoch": 0.3761921582479689,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990643749455045,
      "loss": 6.0047,
      "mean_token_accuracy": 0.1578386515378952,
      "num_tokens": 7931564.0,
      "step": 4260
    },
    {
      "entropy": 6.103243350982666,
      "epoch": 0.37663369833980925,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990615038460667,
      "loss": 5.9993,
      "mean_token_accuracy": 0.1634030133485794,
      "num_tokens": 7940240.0,
      "step": 4265
    },
    {
      "entropy": 6.202239990234375,
      "epoch": 0.3770752384316496,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990586283573916,
      "loss": 5.9641,
      "mean_token_accuracy": 0.15639155656099318,
      "num_tokens": 7949933.0,
      "step": 4270
    },
    {
      "entropy": 6.302160453796387,
      "epoch": 0.3775167785234899,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990557484795355,
      "loss": 6.066,
      "mean_token_accuracy": 0.15802549123764037,
      "num_tokens": 7958561.0,
      "step": 4275
    },
    {
      "entropy": 6.193315505981445,
      "epoch": 0.37795831861533025,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990528642125545,
      "loss": 6.0602,
      "mean_token_accuracy": 0.15381525307893754,
      "num_tokens": 7969549.0,
      "step": 4280
    },
    {
      "entropy": 6.093262386322022,
      "epoch": 0.37839985870717063,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990499755565055,
      "loss": 5.9312,
      "mean_token_accuracy": 0.16241292804479598,
      "num_tokens": 7978141.0,
      "step": 4285
    },
    {
      "entropy": 6.1137525081634525,
      "epoch": 0.37884139879901096,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990470825114448,
      "loss": 6.0138,
      "mean_token_accuracy": 0.163164858520031,
      "num_tokens": 7987543.0,
      "step": 4290
    },
    {
      "entropy": 6.141917085647583,
      "epoch": 0.3792829388908513,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990441850774292,
      "loss": 5.936,
      "mean_token_accuracy": 0.163266521692276,
      "num_tokens": 7996072.0,
      "step": 4295
    },
    {
      "entropy": 6.089475965499878,
      "epoch": 0.3797244789826916,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990412832545155,
      "loss": 5.9484,
      "mean_token_accuracy": 0.16371660232543944,
      "num_tokens": 8005448.0,
      "step": 4300
    },
    {
      "entropy": 6.244927549362183,
      "epoch": 0.38016601907453196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990383770427603,
      "loss": 6.0052,
      "mean_token_accuracy": 0.16179897040128707,
      "num_tokens": 8013845.0,
      "step": 4305
    },
    {
      "entropy": 6.230136156082153,
      "epoch": 0.3806075591663723,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990354664422209,
      "loss": 6.0428,
      "mean_token_accuracy": 0.15771120488643647,
      "num_tokens": 8024048.0,
      "step": 4310
    },
    {
      "entropy": 6.11743311882019,
      "epoch": 0.3810490992582126,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990325514529541,
      "loss": 5.9641,
      "mean_token_accuracy": 0.16579091548919678,
      "num_tokens": 8033242.0,
      "step": 4315
    },
    {
      "entropy": 6.207578182220459,
      "epoch": 0.381490639350053,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990296320750169,
      "loss": 6.0292,
      "mean_token_accuracy": 0.15795896649360658,
      "num_tokens": 8042604.0,
      "step": 4320
    },
    {
      "entropy": 6.148019027709961,
      "epoch": 0.38193217944189334,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990267083084667,
      "loss": 6.0595,
      "mean_token_accuracy": 0.1545356333255768,
      "num_tokens": 8053143.0,
      "step": 4325
    },
    {
      "entropy": 6.230630493164062,
      "epoch": 0.38237371953373367,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990237801533607,
      "loss": 6.0663,
      "mean_token_accuracy": 0.15248100236058235,
      "num_tokens": 8062489.0,
      "step": 4330
    },
    {
      "entropy": 6.216675710678101,
      "epoch": 0.382815259625574,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990208476097562,
      "loss": 5.9393,
      "mean_token_accuracy": 0.17278756201267242,
      "num_tokens": 8071013.0,
      "step": 4335
    },
    {
      "entropy": 6.178141021728516,
      "epoch": 0.38325679971741433,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990179106777109,
      "loss": 6.0457,
      "mean_token_accuracy": 0.1588110461831093,
      "num_tokens": 8080314.0,
      "step": 4340
    },
    {
      "entropy": 6.219320487976074,
      "epoch": 0.38369833980925466,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990149693572819,
      "loss": 5.985,
      "mean_token_accuracy": 0.16360198557376862,
      "num_tokens": 8089903.0,
      "step": 4345
    },
    {
      "entropy": 6.136008405685425,
      "epoch": 0.384139879901095,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990120236485271,
      "loss": 5.9742,
      "mean_token_accuracy": 0.16896202713251113,
      "num_tokens": 8098461.0,
      "step": 4350
    },
    {
      "entropy": 6.132070684432984,
      "epoch": 0.3845814199929354,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990090735515043,
      "loss": 5.9726,
      "mean_token_accuracy": 0.16434564143419267,
      "num_tokens": 8107459.0,
      "step": 4355
    },
    {
      "entropy": 6.221308898925781,
      "epoch": 0.3850229600847757,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499006119066271,
      "loss": 5.9577,
      "mean_token_accuracy": 0.1609977200627327,
      "num_tokens": 8116221.0,
      "step": 4360
    },
    {
      "entropy": 6.133105754852295,
      "epoch": 0.38546450017661604,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990031601928854,
      "loss": 5.8877,
      "mean_token_accuracy": 0.17196132838726044,
      "num_tokens": 8125114.0,
      "step": 4365
    },
    {
      "entropy": 6.116384887695313,
      "epoch": 0.3859060402684564,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990001969314051,
      "loss": 5.9346,
      "mean_token_accuracy": 0.16449921429157258,
      "num_tokens": 8134595.0,
      "step": 4370
    },
    {
      "entropy": 6.134853744506836,
      "epoch": 0.3863475803602967,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989972292818884,
      "loss": 5.9002,
      "mean_token_accuracy": 0.16850025355815887,
      "num_tokens": 8144249.0,
      "step": 4375
    },
    {
      "entropy": 6.060322570800781,
      "epoch": 0.38678912045213704,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989942572443934,
      "loss": 5.8513,
      "mean_token_accuracy": 0.173954139649868,
      "num_tokens": 8153790.0,
      "step": 4380
    },
    {
      "entropy": 6.159169340133667,
      "epoch": 0.38723066054397737,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989912808189784,
      "loss": 5.9859,
      "mean_token_accuracy": 0.16325163543224336,
      "num_tokens": 8162868.0,
      "step": 4385
    },
    {
      "entropy": 6.222345018386841,
      "epoch": 0.38767220063581775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989883000057013,
      "loss": 6.0621,
      "mean_token_accuracy": 0.15853357017040254,
      "num_tokens": 8173018.0,
      "step": 4390
    },
    {
      "entropy": 6.23879714012146,
      "epoch": 0.3881137407276581,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498985314804621,
      "loss": 6.0042,
      "mean_token_accuracy": 0.16005658954381943,
      "num_tokens": 8181743.0,
      "step": 4395
    },
    {
      "entropy": 6.210149335861206,
      "epoch": 0.3885552808194984,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989823252157958,
      "loss": 6.0068,
      "mean_token_accuracy": 0.1549755334854126,
      "num_tokens": 8191378.0,
      "step": 4400
    },
    {
      "entropy": 6.150260400772095,
      "epoch": 0.38899682091133875,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004989793312392841,
      "loss": 5.9446,
      "mean_token_accuracy": 0.16049075722694398,
      "num_tokens": 8200921.0,
      "step": 4405
    },
    {
      "entropy": 6.234221458435059,
      "epoch": 0.3894383610031791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989763328751448,
      "loss": 6.0256,
      "mean_token_accuracy": 0.1567737266421318,
      "num_tokens": 8210507.0,
      "step": 4410
    },
    {
      "entropy": 6.156441926956177,
      "epoch": 0.3898799010950194,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989733301234365,
      "loss": 5.9938,
      "mean_token_accuracy": 0.16142137497663497,
      "num_tokens": 8219583.0,
      "step": 4415
    },
    {
      "entropy": 6.114009857177734,
      "epoch": 0.39032144118685974,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498970322984218,
      "loss": 5.943,
      "mean_token_accuracy": 0.16706262528896332,
      "num_tokens": 8229351.0,
      "step": 4420
    },
    {
      "entropy": 6.104764699935913,
      "epoch": 0.39076298127870013,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989673114575483,
      "loss": 5.8944,
      "mean_token_accuracy": 0.1711021840572357,
      "num_tokens": 8238469.0,
      "step": 4425
    },
    {
      "entropy": 6.246943044662475,
      "epoch": 0.39120452137054046,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989642955434863,
      "loss": 6.0396,
      "mean_token_accuracy": 0.16184400767087936,
      "num_tokens": 8248043.0,
      "step": 4430
    },
    {
      "entropy": 6.170554828643799,
      "epoch": 0.3916460614623808,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004989612752420912,
      "loss": 6.0095,
      "mean_token_accuracy": 0.15891519486904143,
      "num_tokens": 8257372.0,
      "step": 4435
    },
    {
      "entropy": 6.210437393188476,
      "epoch": 0.3920876015542211,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498958250553422,
      "loss": 5.9056,
      "mean_token_accuracy": 0.1694319263100624,
      "num_tokens": 8266954.0,
      "step": 4440
    },
    {
      "entropy": 6.130793666839599,
      "epoch": 0.39252914164606145,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989552214775381,
      "loss": 5.9331,
      "mean_token_accuracy": 0.1702338606119156,
      "num_tokens": 8276100.0,
      "step": 4445
    },
    {
      "entropy": 6.1769571781158445,
      "epoch": 0.3929706817379018,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989521880144988,
      "loss": 5.9633,
      "mean_token_accuracy": 0.16390354484319686,
      "num_tokens": 8285862.0,
      "step": 4450
    },
    {
      "entropy": 6.132815742492676,
      "epoch": 0.3934122218297421,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989491501643635,
      "loss": 5.9235,
      "mean_token_accuracy": 0.16267104893922807,
      "num_tokens": 8294563.0,
      "step": 4455
    },
    {
      "entropy": 6.124179983139038,
      "epoch": 0.3938537619215825,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989461079271916,
      "loss": 5.8533,
      "mean_token_accuracy": 0.1645409345626831,
      "num_tokens": 8303377.0,
      "step": 4460
    },
    {
      "entropy": 6.174312734603882,
      "epoch": 0.39429530201342283,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989430613030429,
      "loss": 6.0304,
      "mean_token_accuracy": 0.16780948638916016,
      "num_tokens": 8312817.0,
      "step": 4465
    },
    {
      "entropy": 6.1341479301452635,
      "epoch": 0.39473684210526316,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498940010291977,
      "loss": 5.9593,
      "mean_token_accuracy": 0.15986717194318772,
      "num_tokens": 8322330.0,
      "step": 4470
    },
    {
      "entropy": 6.239929628372193,
      "epoch": 0.3951783821971035,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989369548940536,
      "loss": 5.9416,
      "mean_token_accuracy": 0.16462937444448472,
      "num_tokens": 8331402.0,
      "step": 4475
    },
    {
      "entropy": 6.190851497650146,
      "epoch": 0.3956199222889438,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989338951093327,
      "loss": 5.9475,
      "mean_token_accuracy": 0.1662589818239212,
      "num_tokens": 8340756.0,
      "step": 4480
    },
    {
      "entropy": 6.100157308578491,
      "epoch": 0.39606146238078416,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989308309378741,
      "loss": 5.9045,
      "mean_token_accuracy": 0.1712504133582115,
      "num_tokens": 8350144.0,
      "step": 4485
    },
    {
      "entropy": 6.117209386825562,
      "epoch": 0.3965030024726245,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989277623797379,
      "loss": 6.0204,
      "mean_token_accuracy": 0.1574763000011444,
      "num_tokens": 8360220.0,
      "step": 4490
    },
    {
      "entropy": 6.167348861694336,
      "epoch": 0.3969445425644649,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989246894349841,
      "loss": 6.0215,
      "mean_token_accuracy": 0.15379244163632394,
      "num_tokens": 8369392.0,
      "step": 4495
    },
    {
      "entropy": 6.224589157104492,
      "epoch": 0.3973860826563052,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989216121036732,
      "loss": 5.9023,
      "mean_token_accuracy": 0.17135855406522751,
      "num_tokens": 8378179.0,
      "step": 4500
    },
    {
      "entropy": 6.206752014160156,
      "epoch": 0.39782762274814554,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989185303858651,
      "loss": 5.9374,
      "mean_token_accuracy": 0.17208127230405806,
      "num_tokens": 8387828.0,
      "step": 4505
    },
    {
      "entropy": 6.109704065322876,
      "epoch": 0.39826916283998587,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989154442816203,
      "loss": 5.9506,
      "mean_token_accuracy": 0.16797782182693483,
      "num_tokens": 8396381.0,
      "step": 4510
    },
    {
      "entropy": 6.101487874984741,
      "epoch": 0.3987107029318262,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004989123537909994,
      "loss": 5.9308,
      "mean_token_accuracy": 0.16774115115404128,
      "num_tokens": 8406775.0,
      "step": 4515
    },
    {
      "entropy": 6.245993900299072,
      "epoch": 0.39915224302366653,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004989092589140629,
      "loss": 6.044,
      "mean_token_accuracy": 0.16268904209136964,
      "num_tokens": 8416818.0,
      "step": 4520
    },
    {
      "entropy": 6.088048505783081,
      "epoch": 0.39959378311550686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989061596508712,
      "loss": 5.8995,
      "mean_token_accuracy": 0.1727058246731758,
      "num_tokens": 8426283.0,
      "step": 4525
    },
    {
      "entropy": 6.167062425613404,
      "epoch": 0.40003532320734725,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989030560014853,
      "loss": 5.9487,
      "mean_token_accuracy": 0.16596201062202454,
      "num_tokens": 8435295.0,
      "step": 4530
    },
    {
      "entropy": 6.1669069766998295,
      "epoch": 0.4004768632991876,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988999479659657,
      "loss": 5.9079,
      "mean_token_accuracy": 0.1674339324235916,
      "num_tokens": 8443882.0,
      "step": 4535
    },
    {
      "entropy": 6.063937187194824,
      "epoch": 0.4009184033910279,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988968355443737,
      "loss": 6.0086,
      "mean_token_accuracy": 0.16480039954185485,
      "num_tokens": 8452678.0,
      "step": 4540
    },
    {
      "entropy": 6.199045896530151,
      "epoch": 0.40135994348286824,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988937187367699,
      "loss": 6.0101,
      "mean_token_accuracy": 0.16075148284435273,
      "num_tokens": 8462315.0,
      "step": 4545
    },
    {
      "entropy": 6.1672642707824705,
      "epoch": 0.4018014835747086,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988905975432154,
      "loss": 5.9514,
      "mean_token_accuracy": 0.16263181939721108,
      "num_tokens": 8471651.0,
      "step": 4550
    },
    {
      "entropy": 6.095368337631226,
      "epoch": 0.4022430236665489,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988874719637715,
      "loss": 5.9158,
      "mean_token_accuracy": 0.16269029527902604,
      "num_tokens": 8480893.0,
      "step": 4555
    },
    {
      "entropy": 6.134046316146851,
      "epoch": 0.40268456375838924,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988843419984994,
      "loss": 5.9983,
      "mean_token_accuracy": 0.16334512382745742,
      "num_tokens": 8490868.0,
      "step": 4560
    },
    {
      "entropy": 6.224040412902832,
      "epoch": 0.4031261038502296,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988812076474604,
      "loss": 5.9823,
      "mean_token_accuracy": 0.15920243188738822,
      "num_tokens": 8500200.0,
      "step": 4565
    },
    {
      "entropy": 6.163804531097412,
      "epoch": 0.40356764394206995,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988780689107158,
      "loss": 6.0234,
      "mean_token_accuracy": 0.1645488053560257,
      "num_tokens": 8509865.0,
      "step": 4570
    },
    {
      "entropy": 6.148987293243408,
      "epoch": 0.4040091840339103,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988749257883271,
      "loss": 5.9895,
      "mean_token_accuracy": 0.16065399572253228,
      "num_tokens": 8519618.0,
      "step": 4575
    },
    {
      "entropy": 6.197585153579712,
      "epoch": 0.4044507241257506,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498871778280356,
      "loss": 5.9597,
      "mean_token_accuracy": 0.15692733228206635,
      "num_tokens": 8529521.0,
      "step": 4580
    },
    {
      "entropy": 6.0856067657470705,
      "epoch": 0.40489226421759095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988686263868641,
      "loss": 5.9586,
      "mean_token_accuracy": 0.1608054056763649,
      "num_tokens": 8539185.0,
      "step": 4585
    },
    {
      "entropy": 6.227382373809815,
      "epoch": 0.4053338043094313,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988654701079131,
      "loss": 5.9555,
      "mean_token_accuracy": 0.16678651422262192,
      "num_tokens": 8548234.0,
      "step": 4590
    },
    {
      "entropy": 6.164825296401977,
      "epoch": 0.4057753444012716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988623094435649,
      "loss": 5.9965,
      "mean_token_accuracy": 0.16253240704536437,
      "num_tokens": 8557001.0,
      "step": 4595
    },
    {
      "entropy": 6.145005512237549,
      "epoch": 0.406216884493112,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988591443938813,
      "loss": 5.9638,
      "mean_token_accuracy": 0.16574816852808,
      "num_tokens": 8565573.0,
      "step": 4600
    },
    {
      "entropy": 6.227346181869507,
      "epoch": 0.4066584245849523,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988559749589244,
      "loss": 5.9931,
      "mean_token_accuracy": 0.1569085881114006,
      "num_tokens": 8575098.0,
      "step": 4605
    },
    {
      "entropy": 6.053036165237427,
      "epoch": 0.40709996467679266,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988528011387563,
      "loss": 5.8795,
      "mean_token_accuracy": 0.17411227375268937,
      "num_tokens": 8584174.0,
      "step": 4610
    },
    {
      "entropy": 6.085353374481201,
      "epoch": 0.407541504768633,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988496229334392,
      "loss": 5.9326,
      "mean_token_accuracy": 0.17281839698553086,
      "num_tokens": 8594124.0,
      "step": 4615
    },
    {
      "entropy": 6.179038333892822,
      "epoch": 0.4079830448604733,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988464403430352,
      "loss": 5.9807,
      "mean_token_accuracy": 0.16023878902196884,
      "num_tokens": 8604085.0,
      "step": 4620
    },
    {
      "entropy": 6.099196481704712,
      "epoch": 0.40842458495231365,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988432533676067,
      "loss": 5.7766,
      "mean_token_accuracy": 0.17047477066516875,
      "num_tokens": 8612539.0,
      "step": 4625
    },
    {
      "entropy": 6.132386207580566,
      "epoch": 0.408866125044154,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988400620072163,
      "loss": 6.0096,
      "mean_token_accuracy": 0.16159712821245192,
      "num_tokens": 8623067.0,
      "step": 4630
    },
    {
      "entropy": 6.086726427078247,
      "epoch": 0.40930766513599437,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988368662619263,
      "loss": 5.8836,
      "mean_token_accuracy": 0.16936124116182327,
      "num_tokens": 8632360.0,
      "step": 4635
    },
    {
      "entropy": 6.207618141174317,
      "epoch": 0.4097492052278347,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988336661317994,
      "loss": 5.9653,
      "mean_token_accuracy": 0.16185452789068222,
      "num_tokens": 8642035.0,
      "step": 4640
    },
    {
      "entropy": 6.11894474029541,
      "epoch": 0.41019074531967503,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988304616168984,
      "loss": 5.9056,
      "mean_token_accuracy": 0.16533755213022233,
      "num_tokens": 8650498.0,
      "step": 4645
    },
    {
      "entropy": 6.0770683765411375,
      "epoch": 0.41063228541151536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988272527172858,
      "loss": 5.9484,
      "mean_token_accuracy": 0.16218477934598924,
      "num_tokens": 8659643.0,
      "step": 4650
    },
    {
      "entropy": 6.172474002838134,
      "epoch": 0.4110738255033557,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988240394330246,
      "loss": 5.9325,
      "mean_token_accuracy": 0.1643434688448906,
      "num_tokens": 8669168.0,
      "step": 4655
    },
    {
      "entropy": 6.165640544891358,
      "epoch": 0.411515365595196,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988208217641778,
      "loss": 5.915,
      "mean_token_accuracy": 0.1641341671347618,
      "num_tokens": 8677750.0,
      "step": 4660
    },
    {
      "entropy": 6.145316410064697,
      "epoch": 0.41195690568703636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988175997108086,
      "loss": 5.969,
      "mean_token_accuracy": 0.16019536852836608,
      "num_tokens": 8686787.0,
      "step": 4665
    },
    {
      "entropy": 6.16266622543335,
      "epoch": 0.41239844577887674,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988143732729797,
      "loss": 5.9462,
      "mean_token_accuracy": 0.16365819573402404,
      "num_tokens": 8695898.0,
      "step": 4670
    },
    {
      "entropy": 6.142560863494873,
      "epoch": 0.4128399858707171,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988111424507546,
      "loss": 5.8782,
      "mean_token_accuracy": 0.16093092262744904,
      "num_tokens": 8705385.0,
      "step": 4675
    },
    {
      "entropy": 6.091668224334716,
      "epoch": 0.4132815259625574,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988079072441964,
      "loss": 5.8979,
      "mean_token_accuracy": 0.17013643980026244,
      "num_tokens": 8715258.0,
      "step": 4680
    },
    {
      "entropy": 6.072128200531006,
      "epoch": 0.41372306605439774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988046676533687,
      "loss": 5.8915,
      "mean_token_accuracy": 0.1633768931031227,
      "num_tokens": 8724106.0,
      "step": 4685
    },
    {
      "entropy": 6.1239495277404785,
      "epoch": 0.41416460614623807,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988014236783347,
      "loss": 5.9888,
      "mean_token_accuracy": 0.15861981511116027,
      "num_tokens": 8732799.0,
      "step": 4690
    },
    {
      "entropy": 6.092017793655396,
      "epoch": 0.4146061462380784,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987981753191582,
      "loss": 5.7835,
      "mean_token_accuracy": 0.17425867170095444,
      "num_tokens": 8741577.0,
      "step": 4695
    },
    {
      "entropy": 6.015551900863647,
      "epoch": 0.41504768632991873,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987949225759027,
      "loss": 5.917,
      "mean_token_accuracy": 0.16846045553684236,
      "num_tokens": 8750914.0,
      "step": 4700
    },
    {
      "entropy": 6.144880294799805,
      "epoch": 0.4154892264217591,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987916654486321,
      "loss": 5.9441,
      "mean_token_accuracy": 0.16389500349760056,
      "num_tokens": 8760186.0,
      "step": 4705
    },
    {
      "entropy": 6.173505640029907,
      "epoch": 0.41593076651359945,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987884039374099,
      "loss": 5.9264,
      "mean_token_accuracy": 0.17249705940485,
      "num_tokens": 8770390.0,
      "step": 4710
    },
    {
      "entropy": 6.059559679031372,
      "epoch": 0.4163723066054398,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987851380423001,
      "loss": 6.0473,
      "mean_token_accuracy": 0.16007738560438156,
      "num_tokens": 8780724.0,
      "step": 4715
    },
    {
      "entropy": 6.144323778152466,
      "epoch": 0.4168138466972801,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987818677633668,
      "loss": 5.874,
      "mean_token_accuracy": 0.17336786091327666,
      "num_tokens": 8789866.0,
      "step": 4720
    },
    {
      "entropy": 6.11774754524231,
      "epoch": 0.41725538678912044,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498778593100674,
      "loss": 5.9546,
      "mean_token_accuracy": 0.17031043618917466,
      "num_tokens": 8799223.0,
      "step": 4725
    },
    {
      "entropy": 6.09026083946228,
      "epoch": 0.4176969268809608,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987753140542857,
      "loss": 5.9994,
      "mean_token_accuracy": 0.16788153648376464,
      "num_tokens": 8809243.0,
      "step": 4730
    },
    {
      "entropy": 6.172619390487671,
      "epoch": 0.4181384669728011,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987720306242664,
      "loss": 5.9339,
      "mean_token_accuracy": 0.16221318989992142,
      "num_tokens": 8818971.0,
      "step": 4735
    },
    {
      "entropy": 6.116506576538086,
      "epoch": 0.4185800070646415,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987687428106803,
      "loss": 5.9056,
      "mean_token_accuracy": 0.1706726387143135,
      "num_tokens": 8827578.0,
      "step": 4740
    },
    {
      "entropy": 6.086148738861084,
      "epoch": 0.4190215471564818,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987654506135917,
      "loss": 5.7624,
      "mean_token_accuracy": 0.1780852720141411,
      "num_tokens": 8836878.0,
      "step": 4745
    },
    {
      "entropy": 6.1360344886779785,
      "epoch": 0.41946308724832215,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987621540330652,
      "loss": 5.9438,
      "mean_token_accuracy": 0.16371864974498748,
      "num_tokens": 8845464.0,
      "step": 4750
    },
    {
      "entropy": 6.075383901596069,
      "epoch": 0.4199046273401625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987588530691653,
      "loss": 5.9211,
      "mean_token_accuracy": 0.1705825924873352,
      "num_tokens": 8854241.0,
      "step": 4755
    },
    {
      "entropy": 6.119292068481445,
      "epoch": 0.4203461674320028,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987555477219569,
      "loss": 5.9657,
      "mean_token_accuracy": 0.16436031460762024,
      "num_tokens": 8864160.0,
      "step": 4760
    },
    {
      "entropy": 6.2072101593017575,
      "epoch": 0.42078770752384315,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004987522379915045,
      "loss": 6.0184,
      "mean_token_accuracy": 0.162161223590374,
      "num_tokens": 8874538.0,
      "step": 4765
    },
    {
      "entropy": 6.113669729232788,
      "epoch": 0.4212292476156835,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498748923877873,
      "loss": 5.9134,
      "mean_token_accuracy": 0.16357507109642028,
      "num_tokens": 8883575.0,
      "step": 4770
    },
    {
      "entropy": 6.119265699386597,
      "epoch": 0.42167078770752386,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987456053811273,
      "loss": 5.9207,
      "mean_token_accuracy": 0.16500757485628129,
      "num_tokens": 8894055.0,
      "step": 4775
    },
    {
      "entropy": 6.063200521469116,
      "epoch": 0.4221123277993642,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987422825013325,
      "loss": 5.8125,
      "mean_token_accuracy": 0.1722533643245697,
      "num_tokens": 8903127.0,
      "step": 4780
    },
    {
      "entropy": 6.110269260406494,
      "epoch": 0.4225538678912045,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987389552385536,
      "loss": 5.9664,
      "mean_token_accuracy": 0.1602931186556816,
      "num_tokens": 8912302.0,
      "step": 4785
    },
    {
      "entropy": 5.99879412651062,
      "epoch": 0.42299540798304486,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987356235928558,
      "loss": 5.8472,
      "mean_token_accuracy": 0.16919654756784439,
      "num_tokens": 8921331.0,
      "step": 4790
    },
    {
      "entropy": 6.202983427047729,
      "epoch": 0.4234369480748852,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987322875643044,
      "loss": 5.9218,
      "mean_token_accuracy": 0.16476156264543534,
      "num_tokens": 8930570.0,
      "step": 4795
    },
    {
      "entropy": 6.180951070785523,
      "epoch": 0.4238784881667255,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004987289471529647,
      "loss": 6.0194,
      "mean_token_accuracy": 0.16260481923818587,
      "num_tokens": 8940099.0,
      "step": 4800
    },
    {
      "entropy": 6.14274377822876,
      "epoch": 0.42432002825856585,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987256023589022,
      "loss": 5.9755,
      "mean_token_accuracy": 0.16113211065530778,
      "num_tokens": 8950951.0,
      "step": 4805
    },
    {
      "entropy": 6.1731914520263675,
      "epoch": 0.42476156835040624,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987222531821824,
      "loss": 6.0233,
      "mean_token_accuracy": 0.16079471707344056,
      "num_tokens": 8961621.0,
      "step": 4810
    },
    {
      "entropy": 6.185737371444702,
      "epoch": 0.42520310844224657,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987188996228709,
      "loss": 5.9439,
      "mean_token_accuracy": 0.16067791879177093,
      "num_tokens": 8970500.0,
      "step": 4815
    },
    {
      "entropy": 6.062719774246216,
      "epoch": 0.4256446485340869,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987155416810334,
      "loss": 5.7909,
      "mean_token_accuracy": 0.17232954502105713,
      "num_tokens": 8979387.0,
      "step": 4820
    },
    {
      "entropy": 6.080672407150269,
      "epoch": 0.42608618862592723,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987121793567356,
      "loss": 5.909,
      "mean_token_accuracy": 0.16659186482429506,
      "num_tokens": 8988819.0,
      "step": 4825
    },
    {
      "entropy": 6.088525199890137,
      "epoch": 0.42652772871776756,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987088126500436,
      "loss": 5.9489,
      "mean_token_accuracy": 0.16552650183439255,
      "num_tokens": 8998712.0,
      "step": 4830
    },
    {
      "entropy": 6.14001579284668,
      "epoch": 0.4269692688096079,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498705441561023,
      "loss": 5.9183,
      "mean_token_accuracy": 0.17174242436885834,
      "num_tokens": 9007829.0,
      "step": 4835
    },
    {
      "entropy": 6.069917440414429,
      "epoch": 0.4274108089014482,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987020660897401,
      "loss": 5.9004,
      "mean_token_accuracy": 0.16427063196897507,
      "num_tokens": 9017756.0,
      "step": 4840
    },
    {
      "entropy": 6.131926679611206,
      "epoch": 0.4278523489932886,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498698686236261,
      "loss": 5.9018,
      "mean_token_accuracy": 0.1690221533179283,
      "num_tokens": 9027003.0,
      "step": 4845
    },
    {
      "entropy": 6.105487108230591,
      "epoch": 0.42829388908512894,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986953020006519,
      "loss": 5.9312,
      "mean_token_accuracy": 0.1644937202334404,
      "num_tokens": 9036458.0,
      "step": 4850
    },
    {
      "entropy": 6.106355619430542,
      "epoch": 0.4287354291769693,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986919133829788,
      "loss": 5.8432,
      "mean_token_accuracy": 0.1670544072985649,
      "num_tokens": 9045206.0,
      "step": 4855
    },
    {
      "entropy": 6.092035913467408,
      "epoch": 0.4291769692688096,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986885203833086,
      "loss": 5.909,
      "mean_token_accuracy": 0.16582580357789994,
      "num_tokens": 9054336.0,
      "step": 4860
    },
    {
      "entropy": 6.112434816360474,
      "epoch": 0.42961850936064994,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986851230017075,
      "loss": 5.8788,
      "mean_token_accuracy": 0.17546966224908828,
      "num_tokens": 9062930.0,
      "step": 4865
    },
    {
      "entropy": 6.112977123260498,
      "epoch": 0.43006004945249027,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986817212382419,
      "loss": 5.9655,
      "mean_token_accuracy": 0.16633554995059968,
      "num_tokens": 9072816.0,
      "step": 4870
    },
    {
      "entropy": 6.078635787963867,
      "epoch": 0.4305015895443306,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986783150929786,
      "loss": 5.8899,
      "mean_token_accuracy": 0.16652937680482865,
      "num_tokens": 9082409.0,
      "step": 4875
    },
    {
      "entropy": 6.105278396606446,
      "epoch": 0.430943129636171,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986749045659845,
      "loss": 5.7842,
      "mean_token_accuracy": 0.17989758551120758,
      "num_tokens": 9091364.0,
      "step": 4880
    },
    {
      "entropy": 6.053403759002686,
      "epoch": 0.4313846697280113,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986714896573261,
      "loss": 5.881,
      "mean_token_accuracy": 0.16614103466272354,
      "num_tokens": 9099993.0,
      "step": 4885
    },
    {
      "entropy": 6.137350177764892,
      "epoch": 0.43182620981985165,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986680703670704,
      "loss": 5.9101,
      "mean_token_accuracy": 0.17212227433919908,
      "num_tokens": 9109391.0,
      "step": 4890
    },
    {
      "entropy": 6.15620985031128,
      "epoch": 0.432267749911692,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986646466952845,
      "loss": 5.9461,
      "mean_token_accuracy": 0.15398568361997605,
      "num_tokens": 9118353.0,
      "step": 4895
    },
    {
      "entropy": 6.025699138641357,
      "epoch": 0.4327092900035323,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986612186420353,
      "loss": 5.9018,
      "mean_token_accuracy": 0.17489442974328995,
      "num_tokens": 9128097.0,
      "step": 4900
    },
    {
      "entropy": 6.173005723953247,
      "epoch": 0.43315083009537264,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986577862073901,
      "loss": 6.0295,
      "mean_token_accuracy": 0.15590395033359528,
      "num_tokens": 9137702.0,
      "step": 4905
    },
    {
      "entropy": 6.167582988739014,
      "epoch": 0.43359237018721297,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986543493914159,
      "loss": 5.9801,
      "mean_token_accuracy": 0.16704000830650328,
      "num_tokens": 9146904.0,
      "step": 4910
    },
    {
      "entropy": 6.07218747138977,
      "epoch": 0.43403391027905336,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986509081941805,
      "loss": 5.8532,
      "mean_token_accuracy": 0.17045014053583146,
      "num_tokens": 9156049.0,
      "step": 4915
    },
    {
      "entropy": 6.144306659698486,
      "epoch": 0.4344754503708937,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986474626157507,
      "loss": 5.9133,
      "mean_token_accuracy": 0.16388186067342758,
      "num_tokens": 9165540.0,
      "step": 4920
    },
    {
      "entropy": 6.131470346450806,
      "epoch": 0.434916990462734,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986440126561945,
      "loss": 5.8939,
      "mean_token_accuracy": 0.1619496151804924,
      "num_tokens": 9174633.0,
      "step": 4925
    },
    {
      "entropy": 6.148479080200195,
      "epoch": 0.43535853055457435,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986405583155792,
      "loss": 5.9153,
      "mean_token_accuracy": 0.1643499806523323,
      "num_tokens": 9184457.0,
      "step": 4930
    },
    {
      "entropy": 6.0906260967254635,
      "epoch": 0.4358000706464147,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986370995939725,
      "loss": 5.9014,
      "mean_token_accuracy": 0.17374202311038972,
      "num_tokens": 9193668.0,
      "step": 4935
    },
    {
      "entropy": 6.1218101978302,
      "epoch": 0.436241610738255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986336364914423,
      "loss": 5.9045,
      "mean_token_accuracy": 0.16776014417409896,
      "num_tokens": 9202410.0,
      "step": 4940
    },
    {
      "entropy": 6.045066165924072,
      "epoch": 0.43668315083009535,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986301690080564,
      "loss": 5.8071,
      "mean_token_accuracy": 0.18001706451177596,
      "num_tokens": 9210564.0,
      "step": 4945
    },
    {
      "entropy": 6.016360425949097,
      "epoch": 0.43712469092193573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986266971438826,
      "loss": 5.865,
      "mean_token_accuracy": 0.17653748393058777,
      "num_tokens": 9219107.0,
      "step": 4950
    },
    {
      "entropy": 6.169685411453247,
      "epoch": 0.43756623101377606,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498623220898989,
      "loss": 6.0098,
      "mean_token_accuracy": 0.16009470224380493,
      "num_tokens": 9228308.0,
      "step": 4955
    },
    {
      "entropy": 6.144251918792724,
      "epoch": 0.4380077711056164,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986197402734436,
      "loss": 5.9042,
      "mean_token_accuracy": 0.16510671824216844,
      "num_tokens": 9237815.0,
      "step": 4960
    },
    {
      "entropy": 6.037957429885864,
      "epoch": 0.4384493111974567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986162552673148,
      "loss": 5.9228,
      "mean_token_accuracy": 0.16932978630065917,
      "num_tokens": 9248258.0,
      "step": 4965
    },
    {
      "entropy": 6.1026135921478275,
      "epoch": 0.43889085128929706,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986127658806706,
      "loss": 5.8416,
      "mean_token_accuracy": 0.17689104974269867,
      "num_tokens": 9258291.0,
      "step": 4970
    },
    {
      "entropy": 6.104307746887207,
      "epoch": 0.4393323913811374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986092721135796,
      "loss": 5.962,
      "mean_token_accuracy": 0.16768431961536406,
      "num_tokens": 9267377.0,
      "step": 4975
    },
    {
      "entropy": 6.118853569030762,
      "epoch": 0.4397739314729777,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986057739661101,
      "loss": 5.9233,
      "mean_token_accuracy": 0.17129827588796614,
      "num_tokens": 9277002.0,
      "step": 4980
    },
    {
      "entropy": 6.14767484664917,
      "epoch": 0.4402154715648181,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986022714383307,
      "loss": 5.9529,
      "mean_token_accuracy": 0.1628987193107605,
      "num_tokens": 9286165.0,
      "step": 4985
    },
    {
      "entropy": 6.145922231674194,
      "epoch": 0.44065701165665844,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985987645303099,
      "loss": 5.9619,
      "mean_token_accuracy": 0.1572280928492546,
      "num_tokens": 9296213.0,
      "step": 4990
    },
    {
      "entropy": 6.088545131683349,
      "epoch": 0.44109855174849877,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985952532421164,
      "loss": 5.88,
      "mean_token_accuracy": 0.16294904947280883,
      "num_tokens": 9305172.0,
      "step": 4995
    },
    {
      "entropy": 6.1045068264007565,
      "epoch": 0.4415400918403391,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985917375738193,
      "loss": 5.9019,
      "mean_token_accuracy": 0.16675015687942504,
      "num_tokens": 9314388.0,
      "step": 5000
    },
    {
      "entropy": 5.991567134857178,
      "epoch": 0.44198163193217943,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004985882175254871,
      "loss": 5.8172,
      "mean_token_accuracy": 0.17150003612041473,
      "num_tokens": 9322934.0,
      "step": 5005
    },
    {
      "entropy": 6.029507493972778,
      "epoch": 0.44242317202401976,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985846930971887,
      "loss": 5.8624,
      "mean_token_accuracy": 0.17287719249725342,
      "num_tokens": 9333023.0,
      "step": 5010
    },
    {
      "entropy": 6.098499917984009,
      "epoch": 0.4428647121158601,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985811642889937,
      "loss": 5.8466,
      "mean_token_accuracy": 0.1704522669315338,
      "num_tokens": 9342260.0,
      "step": 5015
    },
    {
      "entropy": 6.084128379821777,
      "epoch": 0.4433062522077005,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985776311009705,
      "loss": 5.8572,
      "mean_token_accuracy": 0.16558632478117943,
      "num_tokens": 9351764.0,
      "step": 5020
    },
    {
      "entropy": 6.0604455947875975,
      "epoch": 0.4437477922995408,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985740935331888,
      "loss": 5.9015,
      "mean_token_accuracy": 0.1648370534181595,
      "num_tokens": 9361787.0,
      "step": 5025
    },
    {
      "entropy": 6.12150387763977,
      "epoch": 0.44418933239138114,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004985705515857177,
      "loss": 5.9505,
      "mean_token_accuracy": 0.16439836770296096,
      "num_tokens": 9372056.0,
      "step": 5030
    },
    {
      "entropy": 6.211551856994629,
      "epoch": 0.4446308724832215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985670052586268,
      "loss": 5.8981,
      "mean_token_accuracy": 0.17032280415296555,
      "num_tokens": 9379657.0,
      "step": 5035
    },
    {
      "entropy": 6.080951833724976,
      "epoch": 0.4450724125750618,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985634545519853,
      "loss": 5.9016,
      "mean_token_accuracy": 0.17010556757450104,
      "num_tokens": 9389156.0,
      "step": 5040
    },
    {
      "entropy": 6.089050245285034,
      "epoch": 0.44551395266690214,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985598994658629,
      "loss": 5.9219,
      "mean_token_accuracy": 0.16619099378585817,
      "num_tokens": 9397727.0,
      "step": 5045
    },
    {
      "entropy": 6.095297384262085,
      "epoch": 0.44595549275874247,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985563400003291,
      "loss": 5.8962,
      "mean_token_accuracy": 0.16658965200185777,
      "num_tokens": 9407234.0,
      "step": 5050
    },
    {
      "entropy": 6.075574922561645,
      "epoch": 0.44639703285058285,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985527761554539,
      "loss": 5.8331,
      "mean_token_accuracy": 0.17933128774166107,
      "num_tokens": 9416355.0,
      "step": 5055
    },
    {
      "entropy": 6.142249250411988,
      "epoch": 0.4468385729424232,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498549207931307,
      "loss": 5.9869,
      "mean_token_accuracy": 0.15777299106121062,
      "num_tokens": 9425551.0,
      "step": 5060
    },
    {
      "entropy": 6.098491430282593,
      "epoch": 0.4472801130342635,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985456353279581,
      "loss": 5.8825,
      "mean_token_accuracy": 0.17725683003664017,
      "num_tokens": 9434969.0,
      "step": 5065
    },
    {
      "entropy": 6.108680868148804,
      "epoch": 0.44772165312610385,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985420583454774,
      "loss": 5.8731,
      "mean_token_accuracy": 0.1708522692322731,
      "num_tokens": 9444458.0,
      "step": 5070
    },
    {
      "entropy": 6.0295727252960205,
      "epoch": 0.4481631932179442,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004985384769839349,
      "loss": 5.8486,
      "mean_token_accuracy": 0.1756501942873001,
      "num_tokens": 9453591.0,
      "step": 5075
    },
    {
      "entropy": 6.102250814437866,
      "epoch": 0.4486047333097845,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985348912434008,
      "loss": 5.8567,
      "mean_token_accuracy": 0.1688767358660698,
      "num_tokens": 9462461.0,
      "step": 5080
    },
    {
      "entropy": 6.077450513839722,
      "epoch": 0.44904627340162484,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004985313011239452,
      "loss": 5.8506,
      "mean_token_accuracy": 0.16746779084205626,
      "num_tokens": 9472997.0,
      "step": 5085
    },
    {
      "entropy": 6.067001008987427,
      "epoch": 0.4494878134934652,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985277066256388,
      "loss": 5.895,
      "mean_token_accuracy": 0.1737973988056183,
      "num_tokens": 9482797.0,
      "step": 5090
    },
    {
      "entropy": 6.1634057521820065,
      "epoch": 0.44992935358530556,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985241077485515,
      "loss": 5.9469,
      "mean_token_accuracy": 0.16530340909957886,
      "num_tokens": 9492799.0,
      "step": 5095
    },
    {
      "entropy": 6.0241453647613525,
      "epoch": 0.4503708936771459,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985205044927541,
      "loss": 5.8613,
      "mean_token_accuracy": 0.17439087182283403,
      "num_tokens": 9502753.0,
      "step": 5100
    },
    {
      "entropy": 6.087627983093261,
      "epoch": 0.4508124337689862,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985168968583173,
      "loss": 5.8783,
      "mean_token_accuracy": 0.1699455589056015,
      "num_tokens": 9512166.0,
      "step": 5105
    },
    {
      "entropy": 6.131301546096802,
      "epoch": 0.45125397386082655,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985132848453114,
      "loss": 5.9489,
      "mean_token_accuracy": 0.16395388394594193,
      "num_tokens": 9522244.0,
      "step": 5110
    },
    {
      "entropy": 6.043722152709961,
      "epoch": 0.4516955139526669,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985096684538075,
      "loss": 5.8873,
      "mean_token_accuracy": 0.16970051676034928,
      "num_tokens": 9532065.0,
      "step": 5115
    },
    {
      "entropy": 6.138525819778442,
      "epoch": 0.4521370540445072,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985060476838763,
      "loss": 5.9117,
      "mean_token_accuracy": 0.16781965345144273,
      "num_tokens": 9541525.0,
      "step": 5120
    },
    {
      "entropy": 6.036776494979859,
      "epoch": 0.4525785941363476,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985024225355887,
      "loss": 5.8221,
      "mean_token_accuracy": 0.16854558438062667,
      "num_tokens": 9550084.0,
      "step": 5125
    },
    {
      "entropy": 6.112966537475586,
      "epoch": 0.45302013422818793,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984987930090158,
      "loss": 5.8419,
      "mean_token_accuracy": 0.16746425479650498,
      "num_tokens": 9559012.0,
      "step": 5130
    },
    {
      "entropy": 6.033627891540528,
      "epoch": 0.45346167432002826,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984951591042285,
      "loss": 5.8717,
      "mean_token_accuracy": 0.17077850103378295,
      "num_tokens": 9569452.0,
      "step": 5135
    },
    {
      "entropy": 6.043017435073852,
      "epoch": 0.4539032144118686,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004984915208212983,
      "loss": 5.9115,
      "mean_token_accuracy": 0.1732204481959343,
      "num_tokens": 9578985.0,
      "step": 5140
    },
    {
      "entropy": 6.094520568847656,
      "epoch": 0.4543447545037089,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984878781602964,
      "loss": 5.8903,
      "mean_token_accuracy": 0.16360331624746322,
      "num_tokens": 9587834.0,
      "step": 5145
    },
    {
      "entropy": 6.047174692153931,
      "epoch": 0.45478629459554926,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984842311212939,
      "loss": 5.8513,
      "mean_token_accuracy": 0.17342076152563096,
      "num_tokens": 9597607.0,
      "step": 5150
    },
    {
      "entropy": 6.1150837421417235,
      "epoch": 0.4552278346873896,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984805797043625,
      "loss": 5.911,
      "mean_token_accuracy": 0.17308316826820375,
      "num_tokens": 9606445.0,
      "step": 5155
    },
    {
      "entropy": 6.036691188812256,
      "epoch": 0.45566937477923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984769239095736,
      "loss": 5.8281,
      "mean_token_accuracy": 0.17179171591997147,
      "num_tokens": 9615773.0,
      "step": 5160
    },
    {
      "entropy": 6.072480916976929,
      "epoch": 0.4561109148710703,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984732637369989,
      "loss": 5.9052,
      "mean_token_accuracy": 0.16205840557813644,
      "num_tokens": 9624520.0,
      "step": 5165
    },
    {
      "entropy": 6.1490096092224125,
      "epoch": 0.45655245496291064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984695991867099,
      "loss": 5.9217,
      "mean_token_accuracy": 0.16582411974668504,
      "num_tokens": 9633072.0,
      "step": 5170
    },
    {
      "entropy": 6.0410826206207275,
      "epoch": 0.45699399505475097,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984659302587788,
      "loss": 5.8614,
      "mean_token_accuracy": 0.16901664882898332,
      "num_tokens": 9642865.0,
      "step": 5175
    },
    {
      "entropy": 6.068327522277832,
      "epoch": 0.4574355351465913,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498462256953277,
      "loss": 5.8765,
      "mean_token_accuracy": 0.166357684135437,
      "num_tokens": 9651648.0,
      "step": 5180
    },
    {
      "entropy": 6.093669128417969,
      "epoch": 0.45787707523843163,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984585792702767,
      "loss": 5.8613,
      "mean_token_accuracy": 0.16298023313283921,
      "num_tokens": 9661396.0,
      "step": 5185
    },
    {
      "entropy": 6.1114583015441895,
      "epoch": 0.45831861533027196,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004984548972098501,
      "loss": 6.0033,
      "mean_token_accuracy": 0.1688285604119301,
      "num_tokens": 9671211.0,
      "step": 5190
    },
    {
      "entropy": 6.050765609741211,
      "epoch": 0.45876015542211235,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000498451210772069,
      "loss": 5.8952,
      "mean_token_accuracy": 0.17069118618965148,
      "num_tokens": 9681139.0,
      "step": 5195
    },
    {
      "entropy": 6.106564617156982,
      "epoch": 0.4592016955139527,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984475199570058,
      "loss": 5.9413,
      "mean_token_accuracy": 0.1598154827952385,
      "num_tokens": 9690664.0,
      "step": 5200
    },
    {
      "entropy": 6.127637815475464,
      "epoch": 0.459643235605793,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984438247647329,
      "loss": 5.9132,
      "mean_token_accuracy": 0.17412295937538147,
      "num_tokens": 9699966.0,
      "step": 5205
    },
    {
      "entropy": 6.0874875545501705,
      "epoch": 0.46008477569763334,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984401251953223,
      "loss": 5.9863,
      "mean_token_accuracy": 0.1643495514988899,
      "num_tokens": 9709527.0,
      "step": 5210
    },
    {
      "entropy": 6.086588716506958,
      "epoch": 0.4605263157894737,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984364212488469,
      "loss": 5.8825,
      "mean_token_accuracy": 0.16531098261475563,
      "num_tokens": 9719186.0,
      "step": 5215
    },
    {
      "entropy": 6.114411640167236,
      "epoch": 0.460967855881314,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984327129253789,
      "loss": 5.9632,
      "mean_token_accuracy": 0.16362119913101197,
      "num_tokens": 9728261.0,
      "step": 5220
    },
    {
      "entropy": 6.072798728942871,
      "epoch": 0.46140939597315433,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984290002249914,
      "loss": 5.7936,
      "mean_token_accuracy": 0.1745459944009781,
      "num_tokens": 9736731.0,
      "step": 5225
    },
    {
      "entropy": 6.040638494491577,
      "epoch": 0.4618509360649947,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984252831477567,
      "loss": 5.7753,
      "mean_token_accuracy": 0.1785675659775734,
      "num_tokens": 9746540.0,
      "step": 5230
    },
    {
      "entropy": 6.163182353973388,
      "epoch": 0.46229247615683505,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984215616937477,
      "loss": 5.9967,
      "mean_token_accuracy": 0.16362278759479523,
      "num_tokens": 9757322.0,
      "step": 5235
    },
    {
      "entropy": 6.154616355895996,
      "epoch": 0.4627340162486754,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984178358630374,
      "loss": 5.8889,
      "mean_token_accuracy": 0.17241238057613373,
      "num_tokens": 9766742.0,
      "step": 5240
    },
    {
      "entropy": 6.0959173202514645,
      "epoch": 0.4631755563405157,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004984141056556989,
      "loss": 5.8298,
      "mean_token_accuracy": 0.16638700067996978,
      "num_tokens": 9775763.0,
      "step": 5245
    },
    {
      "entropy": 6.058253431320191,
      "epoch": 0.46361709643235605,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984103710718051,
      "loss": 5.8808,
      "mean_token_accuracy": 0.16414181888103485,
      "num_tokens": 9784881.0,
      "step": 5250
    },
    {
      "entropy": 6.118709230422974,
      "epoch": 0.4640586365241964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498406632111429,
      "loss": 5.8922,
      "mean_token_accuracy": 0.16974078863859177,
      "num_tokens": 9794335.0,
      "step": 5255
    },
    {
      "entropy": 6.024377536773682,
      "epoch": 0.4645001766160367,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984028887746443,
      "loss": 5.864,
      "mean_token_accuracy": 0.17572922706604005,
      "num_tokens": 9803063.0,
      "step": 5260
    },
    {
      "entropy": 6.038512229919434,
      "epoch": 0.4649417167078771,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004983991410615239,
      "loss": 5.8786,
      "mean_token_accuracy": 0.1678363248705864,
      "num_tokens": 9812576.0,
      "step": 5265
    },
    {
      "entropy": 6.077331686019898,
      "epoch": 0.4653832567997174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983953889721414,
      "loss": 5.8534,
      "mean_token_accuracy": 0.16549117416143416,
      "num_tokens": 9821910.0,
      "step": 5270
    },
    {
      "entropy": 6.130168151855469,
      "epoch": 0.46582479689155776,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983916325065703,
      "loss": 5.9714,
      "mean_token_accuracy": 0.1623936802148819,
      "num_tokens": 9832327.0,
      "step": 5275
    },
    {
      "entropy": 6.048503112792969,
      "epoch": 0.4662663369833981,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983878716648842,
      "loss": 5.8041,
      "mean_token_accuracy": 0.16593654751777648,
      "num_tokens": 9841624.0,
      "step": 5280
    },
    {
      "entropy": 6.01482138633728,
      "epoch": 0.4667078770752384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983841064471567,
      "loss": 5.8896,
      "mean_token_accuracy": 0.17555419057607652,
      "num_tokens": 9850406.0,
      "step": 5285
    },
    {
      "entropy": 6.088549995422364,
      "epoch": 0.46714941716707875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983803368534617,
      "loss": 5.8442,
      "mean_token_accuracy": 0.17083895802497864,
      "num_tokens": 9858779.0,
      "step": 5290
    },
    {
      "entropy": 6.138989210128784,
      "epoch": 0.4675909572589191,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983765628838728,
      "loss": 5.8882,
      "mean_token_accuracy": 0.16596037447452544,
      "num_tokens": 9867360.0,
      "step": 5295
    },
    {
      "entropy": 6.071683120727539,
      "epoch": 0.46803249735075947,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983727845384641,
      "loss": 5.8576,
      "mean_token_accuracy": 0.17595016062259675,
      "num_tokens": 9877020.0,
      "step": 5300
    },
    {
      "entropy": 6.038946008682251,
      "epoch": 0.4684740374425998,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983690018173096,
      "loss": 5.8377,
      "mean_token_accuracy": 0.17445899695158004,
      "num_tokens": 9885691.0,
      "step": 5305
    },
    {
      "entropy": 6.113630962371826,
      "epoch": 0.46891557753444013,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983652147204834,
      "loss": 5.8691,
      "mean_token_accuracy": 0.16534829139709473,
      "num_tokens": 9895840.0,
      "step": 5310
    },
    {
      "entropy": 6.125788736343384,
      "epoch": 0.46935711762628046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983614232480598,
      "loss": 5.8563,
      "mean_token_accuracy": 0.174843892455101,
      "num_tokens": 9905417.0,
      "step": 5315
    },
    {
      "entropy": 6.151941728591919,
      "epoch": 0.4697986577181208,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983576274001127,
      "loss": 5.9766,
      "mean_token_accuracy": 0.15910756587982178,
      "num_tokens": 9915320.0,
      "step": 5320
    },
    {
      "entropy": 6.140222358703613,
      "epoch": 0.4702401978099611,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498353827176717,
      "loss": 5.956,
      "mean_token_accuracy": 0.1651752457022667,
      "num_tokens": 9924727.0,
      "step": 5325
    },
    {
      "entropy": 6.05612621307373,
      "epoch": 0.4706817379018015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983500225779466,
      "loss": 5.8889,
      "mean_token_accuracy": 0.17028194218873977,
      "num_tokens": 9934383.0,
      "step": 5330
    },
    {
      "entropy": 6.125064897537231,
      "epoch": 0.47112327799364184,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983462136038764,
      "loss": 5.8804,
      "mean_token_accuracy": 0.16605703830718993,
      "num_tokens": 9943489.0,
      "step": 5335
    },
    {
      "entropy": 6.042202615737915,
      "epoch": 0.4715648180854822,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983424002545809,
      "loss": 5.846,
      "mean_token_accuracy": 0.17266706228256226,
      "num_tokens": 9952180.0,
      "step": 5340
    },
    {
      "entropy": 6.07879376411438,
      "epoch": 0.4720063581773225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983385825301348,
      "loss": 5.926,
      "mean_token_accuracy": 0.16514379829168319,
      "num_tokens": 9960899.0,
      "step": 5345
    },
    {
      "entropy": 6.061244869232178,
      "epoch": 0.47244789826916284,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983347604306129,
      "loss": 5.9124,
      "mean_token_accuracy": 0.16564049571752548,
      "num_tokens": 9970477.0,
      "step": 5350
    },
    {
      "entropy": 6.050586128234864,
      "epoch": 0.47288943836100317,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983309339560899,
      "loss": 5.8715,
      "mean_token_accuracy": 0.16953430324792862,
      "num_tokens": 9980037.0,
      "step": 5355
    },
    {
      "entropy": 6.090592956542968,
      "epoch": 0.4733309784528435,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983271031066412,
      "loss": 5.7902,
      "mean_token_accuracy": 0.17550152838230132,
      "num_tokens": 9988966.0,
      "step": 5360
    },
    {
      "entropy": 6.0945734024047855,
      "epoch": 0.4737725185446839,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983232678823414,
      "loss": 5.9255,
      "mean_token_accuracy": 0.16419593691825868,
      "num_tokens": 9998319.0,
      "step": 5365
    },
    {
      "entropy": 6.114677476882934,
      "epoch": 0.4742140586365242,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004983194282832657,
      "loss": 6.0005,
      "mean_token_accuracy": 0.16507325917482377,
      "num_tokens": 10009448.0,
      "step": 5370
    },
    {
      "entropy": 6.051210880279541,
      "epoch": 0.47465559872836455,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983155843094895,
      "loss": 5.8578,
      "mean_token_accuracy": 0.1684434249997139,
      "num_tokens": 10018272.0,
      "step": 5375
    },
    {
      "entropy": 6.078690385818481,
      "epoch": 0.4750971388202049,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983117359610881,
      "loss": 5.9286,
      "mean_token_accuracy": 0.16250931471586227,
      "num_tokens": 10027006.0,
      "step": 5380
    },
    {
      "entropy": 6.160658359527588,
      "epoch": 0.4755386789120452,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983078832381367,
      "loss": 5.9031,
      "mean_token_accuracy": 0.17236902564764023,
      "num_tokens": 10037178.0,
      "step": 5385
    },
    {
      "entropy": 6.1488049030303955,
      "epoch": 0.47598021900388554,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983040261407109,
      "loss": 5.9479,
      "mean_token_accuracy": 0.16149061396718026,
      "num_tokens": 10046413.0,
      "step": 5390
    },
    {
      "entropy": 6.090367126464844,
      "epoch": 0.47642175909572587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983001646688863,
      "loss": 5.9361,
      "mean_token_accuracy": 0.15744809657335282,
      "num_tokens": 10055111.0,
      "step": 5395
    },
    {
      "entropy": 6.0590465545654295,
      "epoch": 0.47686329918756626,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982962988227383,
      "loss": 5.7918,
      "mean_token_accuracy": 0.17789959013462067,
      "num_tokens": 10063835.0,
      "step": 5400
    },
    {
      "entropy": 6.0116602897644045,
      "epoch": 0.4773048392794066,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498292428602343,
      "loss": 5.8641,
      "mean_token_accuracy": 0.1713035821914673,
      "num_tokens": 10073260.0,
      "step": 5405
    },
    {
      "entropy": 6.08366026878357,
      "epoch": 0.4777463793712469,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004982885540077758,
      "loss": 5.8815,
      "mean_token_accuracy": 0.1631835699081421,
      "num_tokens": 10082289.0,
      "step": 5410
    },
    {
      "entropy": 6.067383193969727,
      "epoch": 0.47818791946308725,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982846750391129,
      "loss": 5.876,
      "mean_token_accuracy": 0.17117135971784592,
      "num_tokens": 10091406.0,
      "step": 5415
    },
    {
      "entropy": 6.057268905639648,
      "epoch": 0.4786294595549276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982807916964303,
      "loss": 5.9056,
      "mean_token_accuracy": 0.16902911961078643,
      "num_tokens": 10100526.0,
      "step": 5420
    },
    {
      "entropy": 6.107430458068848,
      "epoch": 0.4790709996467679,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000498276903979804,
      "loss": 5.9653,
      "mean_token_accuracy": 0.16150179803371428,
      "num_tokens": 10110184.0,
      "step": 5425
    },
    {
      "entropy": 6.158305501937866,
      "epoch": 0.47951253973860825,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049827301188931,
      "loss": 5.9008,
      "mean_token_accuracy": 0.16371890306472778,
      "num_tokens": 10118955.0,
      "step": 5430
    },
    {
      "entropy": 6.196906614303589,
      "epoch": 0.47995407983044863,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982691154250247,
      "loss": 6.0096,
      "mean_token_accuracy": 0.1634930580854416,
      "num_tokens": 10128470.0,
      "step": 5435
    },
    {
      "entropy": 6.033510780334472,
      "epoch": 0.48039561992228896,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982652145870245,
      "loss": 5.8396,
      "mean_token_accuracy": 0.16784367561340333,
      "num_tokens": 10137881.0,
      "step": 5440
    },
    {
      "entropy": 6.064654541015625,
      "epoch": 0.4808371600141293,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982613093753856,
      "loss": 5.8812,
      "mean_token_accuracy": 0.1636971041560173,
      "num_tokens": 10146831.0,
      "step": 5445
    },
    {
      "entropy": 6.121941661834716,
      "epoch": 0.4812787001059696,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004982573997901847,
      "loss": 5.778,
      "mean_token_accuracy": 0.18002077043056489,
      "num_tokens": 10157008.0,
      "step": 5450
    },
    {
      "entropy": 6.051258993148804,
      "epoch": 0.48172024019780996,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982534858314982,
      "loss": 5.8199,
      "mean_token_accuracy": 0.17627035230398178,
      "num_tokens": 10165597.0,
      "step": 5455
    },
    {
      "entropy": 6.174269771575927,
      "epoch": 0.4821617802896503,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982495674994031,
      "loss": 5.9183,
      "mean_token_accuracy": 0.1685686632990837,
      "num_tokens": 10175636.0,
      "step": 5460
    },
    {
      "entropy": 6.09394097328186,
      "epoch": 0.4826033203814906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982456447939758,
      "loss": 5.8961,
      "mean_token_accuracy": 0.16258590668439865,
      "num_tokens": 10185726.0,
      "step": 5465
    },
    {
      "entropy": 6.0401012897491455,
      "epoch": 0.483044860473331,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982417177152933,
      "loss": 5.8133,
      "mean_token_accuracy": 0.16894529908895492,
      "num_tokens": 10195065.0,
      "step": 5470
    },
    {
      "entropy": 6.095733976364135,
      "epoch": 0.48348640056517134,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982377862634325,
      "loss": 5.9009,
      "mean_token_accuracy": 0.1667404979467392,
      "num_tokens": 10203810.0,
      "step": 5475
    },
    {
      "entropy": 5.9717050075531,
      "epoch": 0.48392794065701167,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982338504384705,
      "loss": 5.8139,
      "mean_token_accuracy": 0.17824128419160842,
      "num_tokens": 10212896.0,
      "step": 5480
    },
    {
      "entropy": 6.175239419937133,
      "epoch": 0.484369480748852,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982299102404843,
      "loss": 5.8922,
      "mean_token_accuracy": 0.16488052010536194,
      "num_tokens": 10221876.0,
      "step": 5485
    },
    {
      "entropy": 6.081827878952026,
      "epoch": 0.48481102084069233,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498225965669551,
      "loss": 5.8711,
      "mean_token_accuracy": 0.17137998342514038,
      "num_tokens": 10231215.0,
      "step": 5490
    },
    {
      "entropy": 6.036869144439697,
      "epoch": 0.48525256093253266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982220167257482,
      "loss": 5.8959,
      "mean_token_accuracy": 0.17199618071317674,
      "num_tokens": 10239927.0,
      "step": 5495
    },
    {
      "entropy": 6.117075252532959,
      "epoch": 0.485694101024373,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982180634091529,
      "loss": 5.7501,
      "mean_token_accuracy": 0.1804247349500656,
      "num_tokens": 10248494.0,
      "step": 5500
    },
    {
      "entropy": 6.1303871154785154,
      "epoch": 0.4861356411162134,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982141057198427,
      "loss": 5.974,
      "mean_token_accuracy": 0.15832919478416443,
      "num_tokens": 10257795.0,
      "step": 5505
    },
    {
      "entropy": 6.078113412857055,
      "epoch": 0.4865771812080537,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982101436578952,
      "loss": 5.867,
      "mean_token_accuracy": 0.1705743908882141,
      "num_tokens": 10266620.0,
      "step": 5510
    },
    {
      "entropy": 6.0501570224761965,
      "epoch": 0.48701872129989404,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982061772233878,
      "loss": 5.8773,
      "mean_token_accuracy": 0.17452601343393326,
      "num_tokens": 10276493.0,
      "step": 5515
    },
    {
      "entropy": 6.11482605934143,
      "epoch": 0.4874602613917344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982022064163984,
      "loss": 5.9754,
      "mean_token_accuracy": 0.16386598199605942,
      "num_tokens": 10285571.0,
      "step": 5520
    },
    {
      "entropy": 6.036791229248047,
      "epoch": 0.4879018014835747,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981982312370047,
      "loss": 5.728,
      "mean_token_accuracy": 0.18108108192682265,
      "num_tokens": 10294681.0,
      "step": 5525
    },
    {
      "entropy": 6.068901395797729,
      "epoch": 0.48834334157541504,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981942516852847,
      "loss": 5.9089,
      "mean_token_accuracy": 0.1623631462454796,
      "num_tokens": 10304447.0,
      "step": 5530
    },
    {
      "entropy": 6.0416961193084715,
      "epoch": 0.48878488166725537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981902677613161,
      "loss": 5.7797,
      "mean_token_accuracy": 0.18293549567461015,
      "num_tokens": 10312625.0,
      "step": 5535
    },
    {
      "entropy": 5.955431175231934,
      "epoch": 0.48922642175909575,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004981862794651771,
      "loss": 5.8359,
      "mean_token_accuracy": 0.17175227403640747,
      "num_tokens": 10321132.0,
      "step": 5540
    },
    {
      "entropy": 6.075172042846679,
      "epoch": 0.4896679618509361,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981822867969459,
      "loss": 5.8022,
      "mean_token_accuracy": 0.1776400089263916,
      "num_tokens": 10330219.0,
      "step": 5545
    },
    {
      "entropy": 6.119480180740356,
      "epoch": 0.4901095019427764,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981782897567006,
      "loss": 5.9338,
      "mean_token_accuracy": 0.16433441936969756,
      "num_tokens": 10340059.0,
      "step": 5550
    },
    {
      "entropy": 6.06856689453125,
      "epoch": 0.49055104203461675,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981742883445195,
      "loss": 5.9605,
      "mean_token_accuracy": 0.17057447433471679,
      "num_tokens": 10348557.0,
      "step": 5555
    },
    {
      "entropy": 6.151064157485962,
      "epoch": 0.4909925821264571,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498170282560481,
      "loss": 5.9689,
      "mean_token_accuracy": 0.1622586354613304,
      "num_tokens": 10358685.0,
      "step": 5560
    },
    {
      "entropy": 6.120852708816528,
      "epoch": 0.4914341222182974,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981662724046637,
      "loss": 5.8786,
      "mean_token_accuracy": 0.16376424133777617,
      "num_tokens": 10368641.0,
      "step": 5565
    },
    {
      "entropy": 6.121900177001953,
      "epoch": 0.49187566231013774,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498162257877146,
      "loss": 5.894,
      "mean_token_accuracy": 0.16042581349611282,
      "num_tokens": 10378833.0,
      "step": 5570
    },
    {
      "entropy": 6.054622173309326,
      "epoch": 0.4923172024019781,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981582389780065,
      "loss": 5.8973,
      "mean_token_accuracy": 0.1672663450241089,
      "num_tokens": 10387904.0,
      "step": 5575
    },
    {
      "entropy": 6.087834453582763,
      "epoch": 0.49275874249381846,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981542157073241,
      "loss": 5.9866,
      "mean_token_accuracy": 0.15857334136962892,
      "num_tokens": 10397812.0,
      "step": 5580
    },
    {
      "entropy": 6.162728595733642,
      "epoch": 0.4932002825856588,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981501880651775,
      "loss": 5.8247,
      "mean_token_accuracy": 0.1744113340973854,
      "num_tokens": 10407934.0,
      "step": 5585
    },
    {
      "entropy": 6.018968772888184,
      "epoch": 0.4936418226774991,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981461560516457,
      "loss": 5.8327,
      "mean_token_accuracy": 0.17315866947174072,
      "num_tokens": 10417160.0,
      "step": 5590
    },
    {
      "entropy": 6.128939390182495,
      "epoch": 0.49408336276933945,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981421196668075,
      "loss": 5.9131,
      "mean_token_accuracy": 0.16873822510242462,
      "num_tokens": 10427045.0,
      "step": 5595
    },
    {
      "entropy": 6.049278211593628,
      "epoch": 0.4945249028611798,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004981380789107422,
      "loss": 5.7141,
      "mean_token_accuracy": 0.18453337550163268,
      "num_tokens": 10435109.0,
      "step": 5600
    },
    {
      "entropy": 5.994805335998535,
      "epoch": 0.4949664429530201,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981340337835287,
      "loss": 5.805,
      "mean_token_accuracy": 0.17704491913318635,
      "num_tokens": 10445013.0,
      "step": 5605
    },
    {
      "entropy": 6.049015951156616,
      "epoch": 0.4954079830448605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981299842852464,
      "loss": 5.9745,
      "mean_token_accuracy": 0.16148436963558196,
      "num_tokens": 10453983.0,
      "step": 5610
    },
    {
      "entropy": 6.138149166107178,
      "epoch": 0.49584952313670083,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981259304159747,
      "loss": 5.8602,
      "mean_token_accuracy": 0.16689231544733046,
      "num_tokens": 10462829.0,
      "step": 5615
    },
    {
      "entropy": 6.143347215652466,
      "epoch": 0.49629106322854116,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981218721757929,
      "loss": 5.8918,
      "mean_token_accuracy": 0.17119958102703095,
      "num_tokens": 10471419.0,
      "step": 5620
    },
    {
      "entropy": 6.070092868804932,
      "epoch": 0.4967326033203815,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981178095647805,
      "loss": 5.8873,
      "mean_token_accuracy": 0.17170490324497223,
      "num_tokens": 10481398.0,
      "step": 5625
    },
    {
      "entropy": 6.133528375625611,
      "epoch": 0.4971741434122218,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981137425830171,
      "loss": 5.9009,
      "mean_token_accuracy": 0.16134956181049348,
      "num_tokens": 10491255.0,
      "step": 5630
    },
    {
      "entropy": 6.042673635482788,
      "epoch": 0.49761568350406216,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981096712305825,
      "loss": 5.709,
      "mean_token_accuracy": 0.18371844589710234,
      "num_tokens": 10500709.0,
      "step": 5635
    },
    {
      "entropy": 6.061768674850464,
      "epoch": 0.4980572235959025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981055955075561,
      "loss": 5.8603,
      "mean_token_accuracy": 0.1721346378326416,
      "num_tokens": 10509244.0,
      "step": 5640
    },
    {
      "entropy": 6.04087381362915,
      "epoch": 0.4984987636877429,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981015154140181,
      "loss": 5.8222,
      "mean_token_accuracy": 0.1770295187830925,
      "num_tokens": 10517708.0,
      "step": 5645
    },
    {
      "entropy": 6.094776964187622,
      "epoch": 0.4989403037795832,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980974309500483,
      "loss": 5.8637,
      "mean_token_accuracy": 0.16245669573545457,
      "num_tokens": 10526838.0,
      "step": 5650
    },
    {
      "entropy": 6.023749589920044,
      "epoch": 0.49938184387142354,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980933421157267,
      "loss": 5.7928,
      "mean_token_accuracy": 0.17787571400403976,
      "num_tokens": 10536392.0,
      "step": 5655
    },
    {
      "entropy": 6.0897575378417965,
      "epoch": 0.49982338396326387,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980892489111334,
      "loss": 5.9275,
      "mean_token_accuracy": 0.15893392860889435,
      "num_tokens": 10547473.0,
      "step": 5660
    },
    {
      "entropy": 5.997266530990601,
      "epoch": 0.5002649240551043,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980851513363486,
      "loss": 5.8375,
      "mean_token_accuracy": 0.17644696682691574,
      "num_tokens": 10557246.0,
      "step": 5665
    },
    {
      "entropy": 6.020549535751343,
      "epoch": 0.5007064641469445,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980810493914526,
      "loss": 5.8297,
      "mean_token_accuracy": 0.16744026094675063,
      "num_tokens": 10566220.0,
      "step": 5670
    },
    {
      "entropy": 6.05863389968872,
      "epoch": 0.5011480042387849,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980769430765256,
      "loss": 5.7693,
      "mean_token_accuracy": 0.17606477737426757,
      "num_tokens": 10575500.0,
      "step": 5675
    },
    {
      "entropy": 5.952675771713257,
      "epoch": 0.5015895443306252,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980728323916484,
      "loss": 5.789,
      "mean_token_accuracy": 0.17377183884382247,
      "num_tokens": 10584704.0,
      "step": 5680
    },
    {
      "entropy": 6.035444831848144,
      "epoch": 0.5020310844224656,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980687173369009,
      "loss": 5.6894,
      "mean_token_accuracy": 0.17170025557279586,
      "num_tokens": 10592846.0,
      "step": 5685
    },
    {
      "entropy": 6.04313850402832,
      "epoch": 0.5024726245143059,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980645979123644,
      "loss": 5.8489,
      "mean_token_accuracy": 0.17184006422758102,
      "num_tokens": 10601840.0,
      "step": 5690
    },
    {
      "entropy": 6.150271987915039,
      "epoch": 0.5029141646061462,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980604741181192,
      "loss": 5.9232,
      "mean_token_accuracy": 0.16457439810037613,
      "num_tokens": 10611985.0,
      "step": 5695
    },
    {
      "entropy": 6.117373037338257,
      "epoch": 0.5033557046979866,
      "grad_norm": 0.875,
      "learning_rate": 0.0004980563459542461,
      "loss": 5.8973,
      "mean_token_accuracy": 0.16580760031938552,
      "num_tokens": 10621155.0,
      "step": 5700
    },
    {
      "entropy": 6.074395036697387,
      "epoch": 0.5037972447898269,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980522134208261,
      "loss": 5.9224,
      "mean_token_accuracy": 0.16844006031751632,
      "num_tokens": 10631044.0,
      "step": 5705
    },
    {
      "entropy": 6.0484216690063475,
      "epoch": 0.5042387848816673,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980480765179401,
      "loss": 5.7766,
      "mean_token_accuracy": 0.17332978397607804,
      "num_tokens": 10640131.0,
      "step": 5710
    },
    {
      "entropy": 5.980501651763916,
      "epoch": 0.5046803249735076,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980439352456692,
      "loss": 5.7958,
      "mean_token_accuracy": 0.1717753827571869,
      "num_tokens": 10649467.0,
      "step": 5715
    },
    {
      "entropy": 6.049843406677246,
      "epoch": 0.505121865065348,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980397896040944,
      "loss": 5.7844,
      "mean_token_accuracy": 0.17369659394025802,
      "num_tokens": 10658207.0,
      "step": 5720
    },
    {
      "entropy": 6.100121593475341,
      "epoch": 0.5055634051571882,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980356395932969,
      "loss": 5.8682,
      "mean_token_accuracy": 0.1596413418650627,
      "num_tokens": 10667429.0,
      "step": 5725
    },
    {
      "entropy": 6.06561131477356,
      "epoch": 0.5060049452490286,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980314852133581,
      "loss": 5.8473,
      "mean_token_accuracy": 0.17444984316825868,
      "num_tokens": 10677363.0,
      "step": 5730
    },
    {
      "entropy": 6.0741602897644045,
      "epoch": 0.506446485340869,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980273264643594,
      "loss": 5.9012,
      "mean_token_accuracy": 0.16399718075990677,
      "num_tokens": 10687914.0,
      "step": 5735
    },
    {
      "entropy": 6.015149974822998,
      "epoch": 0.5068880254327093,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004980231633463822,
      "loss": 5.8245,
      "mean_token_accuracy": 0.1692004084587097,
      "num_tokens": 10696967.0,
      "step": 5740
    },
    {
      "entropy": 6.024313592910767,
      "epoch": 0.5073295655245497,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980189958595081,
      "loss": 5.8078,
      "mean_token_accuracy": 0.17064205706119537,
      "num_tokens": 10705789.0,
      "step": 5745
    },
    {
      "entropy": 6.096585988998413,
      "epoch": 0.5077711056163899,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980148240038186,
      "loss": 5.8671,
      "mean_token_accuracy": 0.1748196616768837,
      "num_tokens": 10715301.0,
      "step": 5750
    },
    {
      "entropy": 5.99168210029602,
      "epoch": 0.5082126457082303,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004980106477793957,
      "loss": 5.755,
      "mean_token_accuracy": 0.18537474870681764,
      "num_tokens": 10724179.0,
      "step": 5755
    },
    {
      "entropy": 6.017768812179566,
      "epoch": 0.5086541858000706,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980064671863209,
      "loss": 5.85,
      "mean_token_accuracy": 0.16838903427124025,
      "num_tokens": 10733114.0,
      "step": 5760
    },
    {
      "entropy": 6.059729957580567,
      "epoch": 0.509095725891911,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980022822246763,
      "loss": 5.7199,
      "mean_token_accuracy": 0.18333654552698136,
      "num_tokens": 10742185.0,
      "step": 5765
    },
    {
      "entropy": 5.993036603927612,
      "epoch": 0.5095372659837514,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979980928945439,
      "loss": 5.8488,
      "mean_token_accuracy": 0.16945252567529678,
      "num_tokens": 10750883.0,
      "step": 5770
    },
    {
      "entropy": 6.005039548873901,
      "epoch": 0.5099788060755917,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004979938991960056,
      "loss": 5.8705,
      "mean_token_accuracy": 0.16999410539865495,
      "num_tokens": 10761649.0,
      "step": 5775
    },
    {
      "entropy": 6.055377626419068,
      "epoch": 0.510420346167432,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979897011291436,
      "loss": 5.856,
      "mean_token_accuracy": 0.16276892572641372,
      "num_tokens": 10771838.0,
      "step": 5780
    },
    {
      "entropy": 6.1257068634033205,
      "epoch": 0.5108618862592723,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979854986940402,
      "loss": 5.962,
      "mean_token_accuracy": 0.16280137300491332,
      "num_tokens": 10782478.0,
      "step": 5785
    },
    {
      "entropy": 6.121572065353393,
      "epoch": 0.5113034263511127,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979812918907777,
      "loss": 5.8847,
      "mean_token_accuracy": 0.17296641767024995,
      "num_tokens": 10790399.0,
      "step": 5790
    },
    {
      "entropy": 6.111231803894043,
      "epoch": 0.511744966442953,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979770807194385,
      "loss": 5.9335,
      "mean_token_accuracy": 0.16290301978588104,
      "num_tokens": 10800666.0,
      "step": 5795
    },
    {
      "entropy": 6.004605531692505,
      "epoch": 0.5121865065347934,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979728651801051,
      "loss": 5.7811,
      "mean_token_accuracy": 0.17948313802480698,
      "num_tokens": 10809706.0,
      "step": 5800
    },
    {
      "entropy": 6.111881303787231,
      "epoch": 0.5126280466266337,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979686452728602,
      "loss": 5.8864,
      "mean_token_accuracy": 0.16470694690942764,
      "num_tokens": 10819596.0,
      "step": 5805
    },
    {
      "entropy": 6.083578968048096,
      "epoch": 0.513069586718474,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979644209977863,
      "loss": 5.8566,
      "mean_token_accuracy": 0.17016089260578154,
      "num_tokens": 10828223.0,
      "step": 5810
    },
    {
      "entropy": 6.118707704544067,
      "epoch": 0.5135111268103144,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979601923549661,
      "loss": 5.9124,
      "mean_token_accuracy": 0.16581295132637025,
      "num_tokens": 10839627.0,
      "step": 5815
    },
    {
      "entropy": 6.178575229644776,
      "epoch": 0.5139526669021547,
      "grad_norm": 0.875,
      "learning_rate": 0.0004979559593444826,
      "loss": 5.9018,
      "mean_token_accuracy": 0.16683151572942734,
      "num_tokens": 10849550.0,
      "step": 5820
    },
    {
      "entropy": 5.919635677337647,
      "epoch": 0.5143942069939951,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979517219664187,
      "loss": 5.6638,
      "mean_token_accuracy": 0.19036297053098677,
      "num_tokens": 10858055.0,
      "step": 5825
    },
    {
      "entropy": 6.05031909942627,
      "epoch": 0.5148357470858353,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979474802208572,
      "loss": 5.8898,
      "mean_token_accuracy": 0.1738341748714447,
      "num_tokens": 10867367.0,
      "step": 5830
    },
    {
      "entropy": 6.071466112136841,
      "epoch": 0.5152772871776757,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004979432341078816,
      "loss": 5.8398,
      "mean_token_accuracy": 0.1727384090423584,
      "num_tokens": 10876059.0,
      "step": 5835
    },
    {
      "entropy": 6.054691028594971,
      "epoch": 0.5157188272695161,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979389836275746,
      "loss": 5.8077,
      "mean_token_accuracy": 0.17774014472961425,
      "num_tokens": 10885666.0,
      "step": 5840
    },
    {
      "entropy": 6.007467985153198,
      "epoch": 0.5161603673613564,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979347287800198,
      "loss": 5.8236,
      "mean_token_accuracy": 0.17580381482839585,
      "num_tokens": 10894009.0,
      "step": 5845
    },
    {
      "entropy": 6.066354417800904,
      "epoch": 0.5166019074531968,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979304695653005,
      "loss": 5.8096,
      "mean_token_accuracy": 0.17463258653879166,
      "num_tokens": 10903361.0,
      "step": 5850
    },
    {
      "entropy": 6.074882507324219,
      "epoch": 0.5170434475450371,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979262059835001,
      "loss": 5.8628,
      "mean_token_accuracy": 0.17167917490005494,
      "num_tokens": 10912627.0,
      "step": 5855
    },
    {
      "entropy": 6.017969036102295,
      "epoch": 0.5174849876368774,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004979219380347021,
      "loss": 5.7807,
      "mean_token_accuracy": 0.16936404258012772,
      "num_tokens": 10921750.0,
      "step": 5860
    },
    {
      "entropy": 6.085241985321045,
      "epoch": 0.5179265277287177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004979176657189901,
      "loss": 5.8283,
      "mean_token_accuracy": 0.16624539494514465,
      "num_tokens": 10930961.0,
      "step": 5865
    },
    {
      "entropy": 6.061229038238525,
      "epoch": 0.5183680678205581,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979133890364477,
      "loss": 5.8569,
      "mean_token_accuracy": 0.16748277544975282,
      "num_tokens": 10940440.0,
      "step": 5870
    },
    {
      "entropy": 6.062572908401489,
      "epoch": 0.5188096079123985,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497909107987159,
      "loss": 5.8779,
      "mean_token_accuracy": 0.16702806949615479,
      "num_tokens": 10950748.0,
      "step": 5875
    },
    {
      "entropy": 6.129399013519287,
      "epoch": 0.5192511480042388,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979048225712074,
      "loss": 5.8382,
      "mean_token_accuracy": 0.16976635605096818,
      "num_tokens": 10959938.0,
      "step": 5880
    },
    {
      "entropy": 5.939240026473999,
      "epoch": 0.5196926880960792,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979005327886772,
      "loss": 5.768,
      "mean_token_accuracy": 0.17641936540603637,
      "num_tokens": 10969300.0,
      "step": 5885
    },
    {
      "entropy": 5.982491779327392,
      "epoch": 0.5201342281879194,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978962386396525,
      "loss": 5.7434,
      "mean_token_accuracy": 0.18003588318824768,
      "num_tokens": 10978703.0,
      "step": 5890
    },
    {
      "entropy": 6.039892721176147,
      "epoch": 0.5205757682797598,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978919401242171,
      "loss": 5.7808,
      "mean_token_accuracy": 0.1790856659412384,
      "num_tokens": 10987289.0,
      "step": 5895
    },
    {
      "entropy": 6.056543397903442,
      "epoch": 0.5210173083716001,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978876372424553,
      "loss": 5.8445,
      "mean_token_accuracy": 0.17068214416503907,
      "num_tokens": 10997451.0,
      "step": 5900
    },
    {
      "entropy": 6.075278091430664,
      "epoch": 0.5214588484634405,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978833299944515,
      "loss": 5.7703,
      "mean_token_accuracy": 0.17804013192653656,
      "num_tokens": 11006645.0,
      "step": 5905
    },
    {
      "entropy": 6.0533362865448,
      "epoch": 0.5219003885552809,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978790183802901,
      "loss": 5.7925,
      "mean_token_accuracy": 0.1753041386604309,
      "num_tokens": 11015920.0,
      "step": 5910
    },
    {
      "entropy": 5.983188009262085,
      "epoch": 0.5223419286471211,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978747024000554,
      "loss": 5.8059,
      "mean_token_accuracy": 0.17118849009275436,
      "num_tokens": 11025203.0,
      "step": 5915
    },
    {
      "entropy": 5.963877534866333,
      "epoch": 0.5227834687389615,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978703820538321,
      "loss": 5.7633,
      "mean_token_accuracy": 0.17249713093042374,
      "num_tokens": 11034215.0,
      "step": 5920
    },
    {
      "entropy": 6.0389501571655275,
      "epoch": 0.5232250088308018,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978660573417048,
      "loss": 5.8124,
      "mean_token_accuracy": 0.17308756113052368,
      "num_tokens": 11044082.0,
      "step": 5925
    },
    {
      "entropy": 6.165842866897583,
      "epoch": 0.5236665489226422,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497861728263758,
      "loss": 5.9062,
      "mean_token_accuracy": 0.16398635059595107,
      "num_tokens": 11053295.0,
      "step": 5930
    },
    {
      "entropy": 6.09898681640625,
      "epoch": 0.5241080890144825,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978573948200769,
      "loss": 5.901,
      "mean_token_accuracy": 0.1648489698767662,
      "num_tokens": 11062872.0,
      "step": 5935
    },
    {
      "entropy": 6.008897066116333,
      "epoch": 0.5245496291063229,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978530570107461,
      "loss": 5.8847,
      "mean_token_accuracy": 0.17075518518686295,
      "num_tokens": 11071636.0,
      "step": 5940
    },
    {
      "entropy": 6.071086072921753,
      "epoch": 0.5249911691981632,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978487148358506,
      "loss": 5.8843,
      "mean_token_accuracy": 0.16767183989286422,
      "num_tokens": 11080670.0,
      "step": 5945
    },
    {
      "entropy": 6.09007773399353,
      "epoch": 0.5254327092900035,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978443682954756,
      "loss": 5.8605,
      "mean_token_accuracy": 0.17423471063375473,
      "num_tokens": 11090238.0,
      "step": 5950
    },
    {
      "entropy": 6.004768753051758,
      "epoch": 0.5258742493818439,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978400173897061,
      "loss": 5.7656,
      "mean_token_accuracy": 0.1716559961438179,
      "num_tokens": 11098649.0,
      "step": 5955
    },
    {
      "entropy": 6.058194541931153,
      "epoch": 0.5263157894736842,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978356621186275,
      "loss": 5.9224,
      "mean_token_accuracy": 0.16825748533010482,
      "num_tokens": 11108278.0,
      "step": 5960
    },
    {
      "entropy": 6.074383211135864,
      "epoch": 0.5267573295655246,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978313024823249,
      "loss": 5.8765,
      "mean_token_accuracy": 0.16652828305959702,
      "num_tokens": 11116694.0,
      "step": 5965
    },
    {
      "entropy": 6.04515061378479,
      "epoch": 0.5271988696573648,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978269384808839,
      "loss": 5.7895,
      "mean_token_accuracy": 0.17445282489061356,
      "num_tokens": 11125637.0,
      "step": 5970
    },
    {
      "entropy": 6.1411560535430905,
      "epoch": 0.5276404097492052,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978225701143898,
      "loss": 5.8671,
      "mean_token_accuracy": 0.16869430840015412,
      "num_tokens": 11134881.0,
      "step": 5975
    },
    {
      "entropy": 6.048288631439209,
      "epoch": 0.5280819498410456,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978181973829284,
      "loss": 5.8783,
      "mean_token_accuracy": 0.16407397091388704,
      "num_tokens": 11144266.0,
      "step": 5980
    },
    {
      "entropy": 6.086959409713745,
      "epoch": 0.5285234899328859,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978138202865851,
      "loss": 5.8327,
      "mean_token_accuracy": 0.17093676775693895,
      "num_tokens": 11153750.0,
      "step": 5985
    },
    {
      "entropy": 6.042310810089111,
      "epoch": 0.5289650300247263,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978094388254459,
      "loss": 5.8453,
      "mean_token_accuracy": 0.1750756472349167,
      "num_tokens": 11162584.0,
      "step": 5990
    },
    {
      "entropy": 6.061363077163696,
      "epoch": 0.5294065701165666,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978050529995965,
      "loss": 5.7253,
      "mean_token_accuracy": 0.17714042961597443,
      "num_tokens": 11172338.0,
      "step": 5995
    },
    {
      "entropy": 6.030038547515869,
      "epoch": 0.5298481102084069,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978006628091228,
      "loss": 5.8319,
      "mean_token_accuracy": 0.18354289680719377,
      "num_tokens": 11180456.0,
      "step": 6000
    },
    {
      "epoch": 0.5298481102084069,
      "eval_entropy": 5.688593368271064,
      "eval_loss": 5.842279434204102,
      "eval_mean_token_accuracy": 0.17906751633006848,
      "eval_num_tokens": 11180456.0,
      "eval_runtime": 20.2799,
      "eval_samples_per_second": 1741.377,
      "eval_steps_per_second": 217.703,
      "step": 6000
    },
    {
      "entropy": 5.946777248382569,
      "epoch": 0.5302896503002472,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977962682541108,
      "loss": 5.8351,
      "mean_token_accuracy": 0.1670779690146446,
      "num_tokens": 11189169.0,
      "step": 6005
    },
    {
      "entropy": 6.092187023162841,
      "epoch": 0.5307311903920876,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977918693346467,
      "loss": 5.8674,
      "mean_token_accuracy": 0.1699207291007042,
      "num_tokens": 11199609.0,
      "step": 6010
    },
    {
      "entropy": 6.04376449584961,
      "epoch": 0.531172730483928,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977874660508165,
      "loss": 5.814,
      "mean_token_accuracy": 0.17578216642141342,
      "num_tokens": 11208311.0,
      "step": 6015
    },
    {
      "entropy": 5.984195518493652,
      "epoch": 0.5316142705757683,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977830584027065,
      "loss": 5.7826,
      "mean_token_accuracy": 0.17869145721197127,
      "num_tokens": 11217330.0,
      "step": 6020
    },
    {
      "entropy": 5.9880077838897705,
      "epoch": 0.5320558106676087,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977786463904033,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1812600836157799,
      "num_tokens": 11226852.0,
      "step": 6025
    },
    {
      "entropy": 6.08953070640564,
      "epoch": 0.5324973507594489,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977742300139929,
      "loss": 5.8773,
      "mean_token_accuracy": 0.16546263992786409,
      "num_tokens": 11235786.0,
      "step": 6030
    },
    {
      "entropy": 6.143393850326538,
      "epoch": 0.5329388908512893,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977698092735621,
      "loss": 5.8962,
      "mean_token_accuracy": 0.1655338004231453,
      "num_tokens": 11244383.0,
      "step": 6035
    },
    {
      "entropy": 5.949332761764526,
      "epoch": 0.5333804309431296,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977653841691975,
      "loss": 5.749,
      "mean_token_accuracy": 0.17380543500185014,
      "num_tokens": 11253806.0,
      "step": 6040
    },
    {
      "entropy": 5.9758056640625,
      "epoch": 0.53382197103497,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977609547009857,
      "loss": 5.8233,
      "mean_token_accuracy": 0.17357553094625472,
      "num_tokens": 11262970.0,
      "step": 6045
    },
    {
      "entropy": 6.070781660079956,
      "epoch": 0.5342635111268104,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977565208690134,
      "loss": 5.8144,
      "mean_token_accuracy": 0.17423090934753419,
      "num_tokens": 11273514.0,
      "step": 6050
    },
    {
      "entropy": 6.091619920730591,
      "epoch": 0.5347050512186506,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977520826733677,
      "loss": 5.7749,
      "mean_token_accuracy": 0.1750997319817543,
      "num_tokens": 11283119.0,
      "step": 6055
    },
    {
      "entropy": 6.053707599639893,
      "epoch": 0.535146591310491,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977476401141353,
      "loss": 5.8901,
      "mean_token_accuracy": 0.1675879806280136,
      "num_tokens": 11292298.0,
      "step": 6060
    },
    {
      "entropy": 6.032284545898437,
      "epoch": 0.5355881314023313,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004977431931914032,
      "loss": 5.7882,
      "mean_token_accuracy": 0.1713281735777855,
      "num_tokens": 11301212.0,
      "step": 6065
    },
    {
      "entropy": 5.967070531845093,
      "epoch": 0.5360296714941717,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977387419052586,
      "loss": 5.7118,
      "mean_token_accuracy": 0.17758375853300096,
      "num_tokens": 11310322.0,
      "step": 6070
    },
    {
      "entropy": 6.076008415222168,
      "epoch": 0.536471211586012,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497734286255789,
      "loss": 5.8384,
      "mean_token_accuracy": 0.17096902430057526,
      "num_tokens": 11319898.0,
      "step": 6075
    },
    {
      "entropy": 6.017457437515259,
      "epoch": 0.5369127516778524,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977298262430813,
      "loss": 5.75,
      "mean_token_accuracy": 0.18080949932336807,
      "num_tokens": 11328529.0,
      "step": 6080
    },
    {
      "entropy": 5.992965173721314,
      "epoch": 0.5373542917696927,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977253618672229,
      "loss": 5.8156,
      "mean_token_accuracy": 0.17111319154500962,
      "num_tokens": 11337785.0,
      "step": 6085
    },
    {
      "entropy": 6.050319242477417,
      "epoch": 0.537795831861533,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977208931283013,
      "loss": 5.8144,
      "mean_token_accuracy": 0.174222494661808,
      "num_tokens": 11347320.0,
      "step": 6090
    },
    {
      "entropy": 6.005075454711914,
      "epoch": 0.5382373719533734,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977164200264042,
      "loss": 5.7209,
      "mean_token_accuracy": 0.1823834016919136,
      "num_tokens": 11356616.0,
      "step": 6095
    },
    {
      "entropy": 5.980099391937256,
      "epoch": 0.5386789120452137,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977119425616191,
      "loss": 5.8235,
      "mean_token_accuracy": 0.1745764821767807,
      "num_tokens": 11366475.0,
      "step": 6100
    },
    {
      "entropy": 6.0730527400970455,
      "epoch": 0.5391204521370541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977074607340336,
      "loss": 5.8352,
      "mean_token_accuracy": 0.17211222201585769,
      "num_tokens": 11375382.0,
      "step": 6105
    },
    {
      "entropy": 6.132960796356201,
      "epoch": 0.5395619922288943,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004977029745437357,
      "loss": 5.9518,
      "mean_token_accuracy": 0.1627982959151268,
      "num_tokens": 11385575.0,
      "step": 6110
    },
    {
      "entropy": 6.010984420776367,
      "epoch": 0.5400035323207347,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497698483990813,
      "loss": 5.8404,
      "mean_token_accuracy": 0.17140698283910752,
      "num_tokens": 11394987.0,
      "step": 6115
    },
    {
      "entropy": 6.089330720901489,
      "epoch": 0.5404450724125751,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976939890753539,
      "loss": 5.8288,
      "mean_token_accuracy": 0.16876522153615953,
      "num_tokens": 11403979.0,
      "step": 6120
    },
    {
      "entropy": 6.108451557159424,
      "epoch": 0.5408866125044154,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497689489797446,
      "loss": 5.8477,
      "mean_token_accuracy": 0.16119576692581178,
      "num_tokens": 11413345.0,
      "step": 6125
    },
    {
      "entropy": 6.076435947418213,
      "epoch": 0.5413281525962558,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976849861571778,
      "loss": 5.8432,
      "mean_token_accuracy": 0.1716509759426117,
      "num_tokens": 11422729.0,
      "step": 6130
    },
    {
      "entropy": 6.028463649749756,
      "epoch": 0.541769692688096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976804781546373,
      "loss": 5.763,
      "mean_token_accuracy": 0.1803737387061119,
      "num_tokens": 11431939.0,
      "step": 6135
    },
    {
      "entropy": 6.031596899032593,
      "epoch": 0.5422112327799364,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004976759657899129,
      "loss": 5.7948,
      "mean_token_accuracy": 0.1785245269536972,
      "num_tokens": 11442126.0,
      "step": 6140
    },
    {
      "entropy": 6.057561254501342,
      "epoch": 0.5426527728717767,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497671449063093,
      "loss": 5.8268,
      "mean_token_accuracy": 0.1759094476699829,
      "num_tokens": 11451259.0,
      "step": 6145
    },
    {
      "entropy": 5.95423436164856,
      "epoch": 0.5430943129636171,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497666927974266,
      "loss": 5.7451,
      "mean_token_accuracy": 0.17698878198862075,
      "num_tokens": 11459999.0,
      "step": 6150
    },
    {
      "entropy": 6.0173563957214355,
      "epoch": 0.5435358530554575,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976624025235205,
      "loss": 5.7815,
      "mean_token_accuracy": 0.1756902113556862,
      "num_tokens": 11469877.0,
      "step": 6155
    },
    {
      "entropy": 6.088263511657715,
      "epoch": 0.5439773931472978,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004976578727109453,
      "loss": 5.9116,
      "mean_token_accuracy": 0.16403934955596924,
      "num_tokens": 11480187.0,
      "step": 6160
    },
    {
      "entropy": 6.073028135299682,
      "epoch": 0.5444189332391381,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976533385366289,
      "loss": 5.7752,
      "mean_token_accuracy": 0.1727202668786049,
      "num_tokens": 11489614.0,
      "step": 6165
    },
    {
      "entropy": 5.9483479976654055,
      "epoch": 0.5448604733309784,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976488000006602,
      "loss": 5.717,
      "mean_token_accuracy": 0.1730195865035057,
      "num_tokens": 11498185.0,
      "step": 6170
    },
    {
      "entropy": 6.007386016845703,
      "epoch": 0.5453020134228188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976442571031282,
      "loss": 5.8079,
      "mean_token_accuracy": 0.17093525677919388,
      "num_tokens": 11507022.0,
      "step": 6175
    },
    {
      "entropy": 6.087558603286743,
      "epoch": 0.5457435535146591,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976397098441218,
      "loss": 5.8397,
      "mean_token_accuracy": 0.1735919564962387,
      "num_tokens": 11517640.0,
      "step": 6180
    },
    {
      "entropy": 5.978409671783448,
      "epoch": 0.5461850936064995,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004976351582237301,
      "loss": 5.7416,
      "mean_token_accuracy": 0.18195045441389085,
      "num_tokens": 11527917.0,
      "step": 6185
    },
    {
      "entropy": 6.0687487602233885,
      "epoch": 0.5466266336983399,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976306022420423,
      "loss": 5.858,
      "mean_token_accuracy": 0.17338260561227797,
      "num_tokens": 11537056.0,
      "step": 6190
    },
    {
      "entropy": 6.072579669952392,
      "epoch": 0.5470681737901801,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976260418991476,
      "loss": 5.8193,
      "mean_token_accuracy": 0.16629869639873504,
      "num_tokens": 11547215.0,
      "step": 6195
    },
    {
      "entropy": 5.988282966613769,
      "epoch": 0.5475097138820205,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976214771951354,
      "loss": 5.7674,
      "mean_token_accuracy": 0.1776995465159416,
      "num_tokens": 11557593.0,
      "step": 6200
    },
    {
      "entropy": 6.052741670608521,
      "epoch": 0.5479512539738608,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976169081300949,
      "loss": 5.7705,
      "mean_token_accuracy": 0.1805304005742073,
      "num_tokens": 11566144.0,
      "step": 6205
    },
    {
      "entropy": 6.039011716842651,
      "epoch": 0.5483927940657012,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976123347041159,
      "loss": 5.7922,
      "mean_token_accuracy": 0.17619321048259734,
      "num_tokens": 11575478.0,
      "step": 6210
    },
    {
      "entropy": 5.961155939102173,
      "epoch": 0.5488343341575415,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976077569172879,
      "loss": 5.8404,
      "mean_token_accuracy": 0.17435215562582015,
      "num_tokens": 11585348.0,
      "step": 6215
    },
    {
      "entropy": 6.007438325881958,
      "epoch": 0.5492758742493818,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976031747697006,
      "loss": 5.8386,
      "mean_token_accuracy": 0.17225035429000854,
      "num_tokens": 11595320.0,
      "step": 6220
    },
    {
      "entropy": 6.079716682434082,
      "epoch": 0.5497174143412222,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975985882614435,
      "loss": 5.9036,
      "mean_token_accuracy": 0.16361773312091826,
      "num_tokens": 11604489.0,
      "step": 6225
    },
    {
      "entropy": 6.07867465019226,
      "epoch": 0.5501589544330625,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975939973926068,
      "loss": 5.8031,
      "mean_token_accuracy": 0.17128798961639405,
      "num_tokens": 11613902.0,
      "step": 6230
    },
    {
      "entropy": 6.059407663345337,
      "epoch": 0.5506004945249029,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975894021632803,
      "loss": 5.8351,
      "mean_token_accuracy": 0.16993627846240997,
      "num_tokens": 11623795.0,
      "step": 6235
    },
    {
      "entropy": 5.9330525398254395,
      "epoch": 0.5510420346167432,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497584802573554,
      "loss": 5.79,
      "mean_token_accuracy": 0.171390300989151,
      "num_tokens": 11633411.0,
      "step": 6240
    },
    {
      "entropy": 6.034342670440674,
      "epoch": 0.5514835747085836,
      "grad_norm": 1.09375,
      "learning_rate": 0.000497580198623518,
      "loss": 5.8221,
      "mean_token_accuracy": 0.16664735078811646,
      "num_tokens": 11642434.0,
      "step": 6245
    },
    {
      "entropy": 6.069857835769653,
      "epoch": 0.5519251148004238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004975755903132623,
      "loss": 5.7929,
      "mean_token_accuracy": 0.17023066729307174,
      "num_tokens": 11651691.0,
      "step": 6250
    },
    {
      "entropy": 6.001432657241821,
      "epoch": 0.5523666548922642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975709776428776,
      "loss": 5.7683,
      "mean_token_accuracy": 0.17227154970169067,
      "num_tokens": 11660312.0,
      "step": 6255
    },
    {
      "entropy": 6.018537902832032,
      "epoch": 0.5528081949841046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497566360612454,
      "loss": 5.892,
      "mean_token_accuracy": 0.17197067588567733,
      "num_tokens": 11669111.0,
      "step": 6260
    },
    {
      "entropy": 6.12612133026123,
      "epoch": 0.5532497350759449,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497561739222082,
      "loss": 5.8427,
      "mean_token_accuracy": 0.1754047155380249,
      "num_tokens": 11678519.0,
      "step": 6265
    },
    {
      "entropy": 6.004354953765869,
      "epoch": 0.5536912751677853,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975571134718521,
      "loss": 5.7595,
      "mean_token_accuracy": 0.17945225685834884,
      "num_tokens": 11687368.0,
      "step": 6270
    },
    {
      "entropy": 5.966367292404175,
      "epoch": 0.5541328152596255,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975524833618548,
      "loss": 5.8713,
      "mean_token_accuracy": 0.17524985522031783,
      "num_tokens": 11696465.0,
      "step": 6275
    },
    {
      "entropy": 6.0153344631195065,
      "epoch": 0.5545743553514659,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497547848892181,
      "loss": 5.6784,
      "mean_token_accuracy": 0.17638431787490844,
      "num_tokens": 11704763.0,
      "step": 6280
    },
    {
      "entropy": 6.070974493026734,
      "epoch": 0.5550158954433062,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004975432100629215,
      "loss": 5.919,
      "mean_token_accuracy": 0.1660629227757454,
      "num_tokens": 11714866.0,
      "step": 6285
    },
    {
      "entropy": 5.998950624465943,
      "epoch": 0.5554574355351466,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497538566874167,
      "loss": 5.731,
      "mean_token_accuracy": 0.1845327839255333,
      "num_tokens": 11724302.0,
      "step": 6290
    },
    {
      "entropy": 6.0152490615844725,
      "epoch": 0.555898975626987,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004975339193260084,
      "loss": 5.8424,
      "mean_token_accuracy": 0.17546267211437225,
      "num_tokens": 11733585.0,
      "step": 6295
    },
    {
      "entropy": 6.079818201065064,
      "epoch": 0.5563405157188273,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975292674185371,
      "loss": 5.8672,
      "mean_token_accuracy": 0.16521721631288527,
      "num_tokens": 11742345.0,
      "step": 6300
    },
    {
      "entropy": 6.1016449451446535,
      "epoch": 0.5567820558106676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975246111518439,
      "loss": 5.8683,
      "mean_token_accuracy": 0.17068251073360444,
      "num_tokens": 11751972.0,
      "step": 6305
    },
    {
      "entropy": 6.0891687870025635,
      "epoch": 0.5572235959025079,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975199505260202,
      "loss": 5.8729,
      "mean_token_accuracy": 0.16749990731477737,
      "num_tokens": 11760707.0,
      "step": 6310
    },
    {
      "entropy": 6.0758589744567875,
      "epoch": 0.5576651359943483,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975152855411571,
      "loss": 5.8899,
      "mean_token_accuracy": 0.1696454957127571,
      "num_tokens": 11770415.0,
      "step": 6315
    },
    {
      "entropy": 6.087446165084839,
      "epoch": 0.5581066760861886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975106161973462,
      "loss": 5.7965,
      "mean_token_accuracy": 0.17424398958683013,
      "num_tokens": 11779255.0,
      "step": 6320
    },
    {
      "entropy": 6.002763175964356,
      "epoch": 0.558548216178029,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497505942494679,
      "loss": 5.7578,
      "mean_token_accuracy": 0.17807706892490388,
      "num_tokens": 11788165.0,
      "step": 6325
    },
    {
      "entropy": 5.94511981010437,
      "epoch": 0.5589897562698694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975012644332467,
      "loss": 5.8447,
      "mean_token_accuracy": 0.1774800166487694,
      "num_tokens": 11797270.0,
      "step": 6330
    },
    {
      "entropy": 6.063098573684693,
      "epoch": 0.5594312963617096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974965820131413,
      "loss": 5.8557,
      "mean_token_accuracy": 0.1668801188468933,
      "num_tokens": 11805892.0,
      "step": 6335
    },
    {
      "entropy": 6.08272910118103,
      "epoch": 0.55987283645355,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004974918952344544,
      "loss": 5.7474,
      "mean_token_accuracy": 0.1766267791390419,
      "num_tokens": 11815189.0,
      "step": 6340
    },
    {
      "entropy": 6.053334093093872,
      "epoch": 0.5603143765453903,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974872040972777,
      "loss": 5.7778,
      "mean_token_accuracy": 0.1768811896443367,
      "num_tokens": 11824371.0,
      "step": 6345
    },
    {
      "entropy": 5.971737098693848,
      "epoch": 0.5607559166372307,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974825086017033,
      "loss": 5.7422,
      "mean_token_accuracy": 0.1752060055732727,
      "num_tokens": 11834455.0,
      "step": 6350
    },
    {
      "entropy": 6.054451322555542,
      "epoch": 0.561197456729071,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497477808747823,
      "loss": 5.8087,
      "mean_token_accuracy": 0.17288438528776168,
      "num_tokens": 11843197.0,
      "step": 6355
    },
    {
      "entropy": 6.034222650527954,
      "epoch": 0.5616389968209113,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974731045357292,
      "loss": 5.7593,
      "mean_token_accuracy": 0.17509490251541138,
      "num_tokens": 11852232.0,
      "step": 6360
    },
    {
      "entropy": 5.974319934844971,
      "epoch": 0.5620805369127517,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974683959655136,
      "loss": 5.782,
      "mean_token_accuracy": 0.16567135602235794,
      "num_tokens": 11860685.0,
      "step": 6365
    },
    {
      "entropy": 5.993752908706665,
      "epoch": 0.562522077004592,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974636830372686,
      "loss": 5.792,
      "mean_token_accuracy": 0.17670380622148513,
      "num_tokens": 11871063.0,
      "step": 6370
    },
    {
      "entropy": 6.040469551086426,
      "epoch": 0.5629636170964324,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974589657510866,
      "loss": 5.8439,
      "mean_token_accuracy": 0.17252148538827897,
      "num_tokens": 11881118.0,
      "step": 6375
    },
    {
      "entropy": 6.0495795726776125,
      "epoch": 0.5634051571882727,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974542441070601,
      "loss": 5.8352,
      "mean_token_accuracy": 0.16615308672189713,
      "num_tokens": 11891227.0,
      "step": 6380
    },
    {
      "entropy": 5.970382785797119,
      "epoch": 0.563846697280113,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974495181052814,
      "loss": 5.753,
      "mean_token_accuracy": 0.17418700754642485,
      "num_tokens": 11900800.0,
      "step": 6385
    },
    {
      "entropy": 6.022345733642578,
      "epoch": 0.5642882373719533,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004974447877458433,
      "loss": 5.7314,
      "mean_token_accuracy": 0.18001427948474885,
      "num_tokens": 11909199.0,
      "step": 6390
    },
    {
      "entropy": 6.090434885025024,
      "epoch": 0.5647297774637937,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974400530288381,
      "loss": 5.8083,
      "mean_token_accuracy": 0.176697014272213,
      "num_tokens": 11918074.0,
      "step": 6395
    },
    {
      "entropy": 6.029108762741089,
      "epoch": 0.5651713175556341,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974353139543589,
      "loss": 5.828,
      "mean_token_accuracy": 0.1693344756960869,
      "num_tokens": 11927047.0,
      "step": 6400
    },
    {
      "entropy": 6.0067192077636715,
      "epoch": 0.5656128576474744,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974305705224983,
      "loss": 5.7,
      "mean_token_accuracy": 0.1815037176012993,
      "num_tokens": 11936163.0,
      "step": 6405
    },
    {
      "entropy": 5.9384850025177,
      "epoch": 0.5660543977393148,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004974258227333493,
      "loss": 5.6714,
      "mean_token_accuracy": 0.17884492874145508,
      "num_tokens": 11946623.0,
      "step": 6410
    },
    {
      "entropy": 5.953594541549682,
      "epoch": 0.566495937831155,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004974210705870052,
      "loss": 5.7216,
      "mean_token_accuracy": 0.1824912503361702,
      "num_tokens": 11955926.0,
      "step": 6415
    },
    {
      "entropy": 6.046787214279175,
      "epoch": 0.5669374779229954,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974163140835586,
      "loss": 5.7873,
      "mean_token_accuracy": 0.17135222405195236,
      "num_tokens": 11964250.0,
      "step": 6420
    },
    {
      "entropy": 5.931289148330689,
      "epoch": 0.5673790180148357,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974115532231029,
      "loss": 5.6862,
      "mean_token_accuracy": 0.18259892463684083,
      "num_tokens": 11973230.0,
      "step": 6425
    },
    {
      "entropy": 5.99631404876709,
      "epoch": 0.5678205581066761,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974067880057315,
      "loss": 5.7864,
      "mean_token_accuracy": 0.17200641185045243,
      "num_tokens": 11982079.0,
      "step": 6430
    },
    {
      "entropy": 6.0347252368927,
      "epoch": 0.5682620981985165,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974020184315375,
      "loss": 5.7929,
      "mean_token_accuracy": 0.1832293152809143,
      "num_tokens": 11991534.0,
      "step": 6435
    },
    {
      "entropy": 6.088953924179077,
      "epoch": 0.5687036382903567,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973972445006144,
      "loss": 5.8062,
      "mean_token_accuracy": 0.17041926383972167,
      "num_tokens": 12001068.0,
      "step": 6440
    },
    {
      "entropy": 6.0781982898712155,
      "epoch": 0.5691451783821971,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004973924662130559,
      "loss": 5.7971,
      "mean_token_accuracy": 0.16880759745836257,
      "num_tokens": 12009941.0,
      "step": 6445
    },
    {
      "entropy": 5.962063884735107,
      "epoch": 0.5695867184740374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973876835689554,
      "loss": 5.6812,
      "mean_token_accuracy": 0.18154624998569488,
      "num_tokens": 12019025.0,
      "step": 6450
    },
    {
      "entropy": 6.012698221206665,
      "epoch": 0.5700282585658778,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004973828965684066,
      "loss": 5.7978,
      "mean_token_accuracy": 0.17401303797960282,
      "num_tokens": 12028938.0,
      "step": 6455
    },
    {
      "entropy": 6.011123609542847,
      "epoch": 0.5704697986577181,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973781052115035,
      "loss": 5.9167,
      "mean_token_accuracy": 0.164264777302742,
      "num_tokens": 12040474.0,
      "step": 6460
    },
    {
      "entropy": 5.9840068340301515,
      "epoch": 0.5709113387495585,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973733094983396,
      "loss": 5.6617,
      "mean_token_accuracy": 0.17362459599971772,
      "num_tokens": 12049370.0,
      "step": 6465
    },
    {
      "entropy": 6.0229172706604,
      "epoch": 0.5713528788413988,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004973685094290092,
      "loss": 5.7314,
      "mean_token_accuracy": 0.17876823991537094,
      "num_tokens": 12058039.0,
      "step": 6470
    },
    {
      "entropy": 5.967266464233399,
      "epoch": 0.5717944189332391,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497363705003606,
      "loss": 5.6783,
      "mean_token_accuracy": 0.18561322689056398,
      "num_tokens": 12067560.0,
      "step": 6475
    },
    {
      "entropy": 6.015367841720581,
      "epoch": 0.5722359590250795,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973588962222244,
      "loss": 5.8639,
      "mean_token_accuracy": 0.17167299389839172,
      "num_tokens": 12076817.0,
      "step": 6480
    },
    {
      "entropy": 5.999601411819458,
      "epoch": 0.5726774991169198,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973540830849585,
      "loss": 5.8671,
      "mean_token_accuracy": 0.17174146473407745,
      "num_tokens": 12086140.0,
      "step": 6485
    },
    {
      "entropy": 5.9419807434082035,
      "epoch": 0.5731190392087602,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973492655919025,
      "loss": 5.7178,
      "mean_token_accuracy": 0.18064942955970764,
      "num_tokens": 12095473.0,
      "step": 6490
    },
    {
      "entropy": 5.963439083099365,
      "epoch": 0.5735605793006004,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004973444437431508,
      "loss": 5.7493,
      "mean_token_accuracy": 0.17527091056108474,
      "num_tokens": 12104712.0,
      "step": 6495
    },
    {
      "entropy": 5.9865974426269535,
      "epoch": 0.5740021193924408,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973396175387981,
      "loss": 5.7261,
      "mean_token_accuracy": 0.1762429505586624,
      "num_tokens": 12113025.0,
      "step": 6500
    },
    {
      "entropy": 6.051536989212036,
      "epoch": 0.5744436594842812,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973347869789385,
      "loss": 5.8458,
      "mean_token_accuracy": 0.1719760075211525,
      "num_tokens": 12122447.0,
      "step": 6505
    },
    {
      "entropy": 5.929996204376221,
      "epoch": 0.5748851995761215,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973299520636671,
      "loss": 5.7498,
      "mean_token_accuracy": 0.17197403013706208,
      "num_tokens": 12131745.0,
      "step": 6510
    },
    {
      "entropy": 5.991893148422241,
      "epoch": 0.5753267396679619,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973251127930782,
      "loss": 5.7611,
      "mean_token_accuracy": 0.1739878848195076,
      "num_tokens": 12140890.0,
      "step": 6515
    },
    {
      "entropy": 5.998748302459717,
      "epoch": 0.5757682797598022,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004973202691672669,
      "loss": 5.7371,
      "mean_token_accuracy": 0.17626408636569976,
      "num_tokens": 12150068.0,
      "step": 6520
    },
    {
      "entropy": 5.92770471572876,
      "epoch": 0.5762098198516425,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004973154211863279,
      "loss": 5.8096,
      "mean_token_accuracy": 0.16779339015483857,
      "num_tokens": 12158135.0,
      "step": 6525
    },
    {
      "entropy": 6.005129003524781,
      "epoch": 0.5766513599434828,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004973105688503563,
      "loss": 5.8173,
      "mean_token_accuracy": 0.17955590337514876,
      "num_tokens": 12166549.0,
      "step": 6530
    },
    {
      "entropy": 6.061457347869873,
      "epoch": 0.5770929000353232,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497305712159447,
      "loss": 5.8989,
      "mean_token_accuracy": 0.17019106447696686,
      "num_tokens": 12175672.0,
      "step": 6535
    },
    {
      "entropy": 6.008888483047485,
      "epoch": 0.5775344401271636,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973008511136954,
      "loss": 5.7348,
      "mean_token_accuracy": 0.17478324919939042,
      "num_tokens": 12184577.0,
      "step": 6540
    },
    {
      "entropy": 5.9324697971344,
      "epoch": 0.5779759802190039,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972959857131963,
      "loss": 5.764,
      "mean_token_accuracy": 0.18289703279733657,
      "num_tokens": 12194690.0,
      "step": 6545
    },
    {
      "entropy": 6.066585731506348,
      "epoch": 0.5784175203108443,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972911159580454,
      "loss": 5.7972,
      "mean_token_accuracy": 0.1771732747554779,
      "num_tokens": 12203813.0,
      "step": 6550
    },
    {
      "entropy": 6.0006263256073,
      "epoch": 0.5788590604026845,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497286241848338,
      "loss": 5.7969,
      "mean_token_accuracy": 0.17684610635042192,
      "num_tokens": 12212736.0,
      "step": 6555
    },
    {
      "entropy": 5.925478982925415,
      "epoch": 0.5793006004945249,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004972813633841696,
      "loss": 5.707,
      "mean_token_accuracy": 0.17733784765005112,
      "num_tokens": 12221605.0,
      "step": 6560
    },
    {
      "entropy": 6.0453064918518065,
      "epoch": 0.5797421405863652,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972764805656355,
      "loss": 5.7708,
      "mean_token_accuracy": 0.17744192779064177,
      "num_tokens": 12231043.0,
      "step": 6565
    },
    {
      "entropy": 5.922590351104736,
      "epoch": 0.5801836806782056,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004972715933928318,
      "loss": 5.8104,
      "mean_token_accuracy": 0.16921428591012955,
      "num_tokens": 12241309.0,
      "step": 6570
    },
    {
      "entropy": 5.964745235443115,
      "epoch": 0.580625220770046,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972667018658539,
      "loss": 5.7758,
      "mean_token_accuracy": 0.17969550639390947,
      "num_tokens": 12250711.0,
      "step": 6575
    },
    {
      "entropy": 5.952535247802734,
      "epoch": 0.5810667608618862,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972618059847976,
      "loss": 5.7,
      "mean_token_accuracy": 0.17868800908327104,
      "num_tokens": 12259317.0,
      "step": 6580
    },
    {
      "entropy": 6.021958112716675,
      "epoch": 0.5815083009537266,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972569057497591,
      "loss": 5.7558,
      "mean_token_accuracy": 0.1752405732870102,
      "num_tokens": 12268342.0,
      "step": 6585
    },
    {
      "entropy": 5.952702522277832,
      "epoch": 0.5819498410455669,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972520011608341,
      "loss": 5.777,
      "mean_token_accuracy": 0.177975133061409,
      "num_tokens": 12277592.0,
      "step": 6590
    },
    {
      "entropy": 5.943783950805664,
      "epoch": 0.5823913811374073,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972470922181189,
      "loss": 5.7223,
      "mean_token_accuracy": 0.1808294400572777,
      "num_tokens": 12286779.0,
      "step": 6595
    },
    {
      "entropy": 6.030092048645019,
      "epoch": 0.5828329212292476,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972421789217094,
      "loss": 5.8528,
      "mean_token_accuracy": 0.1690855711698532,
      "num_tokens": 12297887.0,
      "step": 6600
    },
    {
      "entropy": 6.073009967803955,
      "epoch": 0.583274461321088,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972372612717022,
      "loss": 5.774,
      "mean_token_accuracy": 0.17609166949987412,
      "num_tokens": 12306690.0,
      "step": 6605
    },
    {
      "entropy": 5.981056451797485,
      "epoch": 0.5837160014129283,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972323392681934,
      "loss": 5.7697,
      "mean_token_accuracy": 0.17589016407728195,
      "num_tokens": 12315311.0,
      "step": 6610
    },
    {
      "entropy": 5.958033752441406,
      "epoch": 0.5841575415047686,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972274129112794,
      "loss": 5.7556,
      "mean_token_accuracy": 0.17343640476465225,
      "num_tokens": 12324230.0,
      "step": 6615
    },
    {
      "entropy": 6.05591344833374,
      "epoch": 0.584599081596609,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972224822010567,
      "loss": 5.921,
      "mean_token_accuracy": 0.16143234670162201,
      "num_tokens": 12334122.0,
      "step": 6620
    },
    {
      "entropy": 6.011671495437622,
      "epoch": 0.5850406216884493,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972175471376222,
      "loss": 5.7223,
      "mean_token_accuracy": 0.1806204304099083,
      "num_tokens": 12342221.0,
      "step": 6625
    },
    {
      "entropy": 5.95959620475769,
      "epoch": 0.5854821617802897,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004972126077210722,
      "loss": 5.7137,
      "mean_token_accuracy": 0.1848437637090683,
      "num_tokens": 12350195.0,
      "step": 6630
    },
    {
      "entropy": 5.876103258132934,
      "epoch": 0.5859237018721299,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972076639515035,
      "loss": 5.6996,
      "mean_token_accuracy": 0.18036881238222122,
      "num_tokens": 12359679.0,
      "step": 6635
    },
    {
      "entropy": 5.998584890365601,
      "epoch": 0.5863652419639703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972027158290131,
      "loss": 5.6909,
      "mean_token_accuracy": 0.19083465337753297,
      "num_tokens": 12367822.0,
      "step": 6640
    },
    {
      "entropy": 5.983862590789795,
      "epoch": 0.5868067820558107,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497197763353698,
      "loss": 5.7227,
      "mean_token_accuracy": 0.17392620593309402,
      "num_tokens": 12377198.0,
      "step": 6645
    },
    {
      "entropy": 6.014743137359619,
      "epoch": 0.587248322147651,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497192806525655,
      "loss": 5.8379,
      "mean_token_accuracy": 0.1753952294588089,
      "num_tokens": 12387655.0,
      "step": 6650
    },
    {
      "entropy": 6.029316854476929,
      "epoch": 0.5876898622394914,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004971878453449811,
      "loss": 5.8679,
      "mean_token_accuracy": 0.16357576251029968,
      "num_tokens": 12398143.0,
      "step": 6655
    },
    {
      "entropy": 6.042935037612915,
      "epoch": 0.5881314023313317,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497182879811774,
      "loss": 5.8571,
      "mean_token_accuracy": 0.17424555867910385,
      "num_tokens": 12407355.0,
      "step": 6660
    },
    {
      "entropy": 6.078473615646362,
      "epoch": 0.588572942423172,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004971779099261304,
      "loss": 5.8785,
      "mean_token_accuracy": 0.16561560779809953,
      "num_tokens": 12417526.0,
      "step": 6665
    },
    {
      "entropy": 6.0052238464355465,
      "epoch": 0.5890144825150123,
      "grad_norm": 0.796875,
      "learning_rate": 0.000497172935688148,
      "loss": 5.7543,
      "mean_token_accuracy": 0.1776171773672104,
      "num_tokens": 12427989.0,
      "step": 6670
    },
    {
      "entropy": 6.058671569824218,
      "epoch": 0.5894560226068527,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971679570979242,
      "loss": 5.8937,
      "mean_token_accuracy": 0.16729942411184312,
      "num_tokens": 12437934.0,
      "step": 6675
    },
    {
      "entropy": 6.0393280506134035,
      "epoch": 0.5898975626986931,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004971629741555565,
      "loss": 5.7773,
      "mean_token_accuracy": 0.17749162316322326,
      "num_tokens": 12447466.0,
      "step": 6680
    },
    {
      "entropy": 5.967742156982422,
      "epoch": 0.5903391027905334,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971579868611423,
      "loss": 5.7302,
      "mean_token_accuracy": 0.18247907161712645,
      "num_tokens": 12456326.0,
      "step": 6685
    },
    {
      "entropy": 5.941596984863281,
      "epoch": 0.5907806428823738,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971529952147797,
      "loss": 5.7116,
      "mean_token_accuracy": 0.182989501953125,
      "num_tokens": 12466238.0,
      "step": 6690
    },
    {
      "entropy": 5.978118085861206,
      "epoch": 0.591222182974214,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497147999216566,
      "loss": 5.6926,
      "mean_token_accuracy": 0.18029287606477737,
      "num_tokens": 12475464.0,
      "step": 6695
    },
    {
      "entropy": 6.001824951171875,
      "epoch": 0.5916637230660544,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004971429988665996,
      "loss": 5.7947,
      "mean_token_accuracy": 0.17166142761707306,
      "num_tokens": 12484999.0,
      "step": 6700
    },
    {
      "entropy": 6.090221834182739,
      "epoch": 0.5921052631578947,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497137994164978,
      "loss": 5.7709,
      "mean_token_accuracy": 0.17268445789813996,
      "num_tokens": 12493992.0,
      "step": 6705
    },
    {
      "entropy": 6.013251686096192,
      "epoch": 0.5925468032497351,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004971329851117996,
      "loss": 5.8351,
      "mean_token_accuracy": 0.17588648796081544,
      "num_tokens": 12503611.0,
      "step": 6710
    },
    {
      "entropy": 5.9967645645141605,
      "epoch": 0.5929883433415755,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971279717071622,
      "loss": 5.76,
      "mean_token_accuracy": 0.17788553833961487,
      "num_tokens": 12512482.0,
      "step": 6715
    },
    {
      "entropy": 6.058877182006836,
      "epoch": 0.5934298834334157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971229539511643,
      "loss": 5.7492,
      "mean_token_accuracy": 0.17468193024396897,
      "num_tokens": 12520514.0,
      "step": 6720
    },
    {
      "entropy": 6.011536073684693,
      "epoch": 0.5938714235252561,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497117931843904,
      "loss": 5.7564,
      "mean_token_accuracy": 0.16796766519546508,
      "num_tokens": 12529052.0,
      "step": 6725
    },
    {
      "entropy": 5.978758811950684,
      "epoch": 0.5943129636170964,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004971129053854797,
      "loss": 5.8184,
      "mean_token_accuracy": 0.16544206142425538,
      "num_tokens": 12538420.0,
      "step": 6730
    },
    {
      "entropy": 5.9646100997924805,
      "epoch": 0.5947545037089368,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049710787457599,
      "loss": 5.6884,
      "mean_token_accuracy": 0.18099527657032013,
      "num_tokens": 12547283.0,
      "step": 6735
    },
    {
      "entropy": 6.063982677459717,
      "epoch": 0.5951960438007771,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004971028394155332,
      "loss": 5.7854,
      "mean_token_accuracy": 0.1724291115999222,
      "num_tokens": 12556532.0,
      "step": 6740
    },
    {
      "entropy": 5.95719165802002,
      "epoch": 0.5956375838926175,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970977999042082,
      "loss": 5.7785,
      "mean_token_accuracy": 0.16847883462905883,
      "num_tokens": 12566242.0,
      "step": 6745
    },
    {
      "entropy": 5.932070398330689,
      "epoch": 0.5960791239844578,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970927560421137,
      "loss": 5.7595,
      "mean_token_accuracy": 0.1758987307548523,
      "num_tokens": 12575031.0,
      "step": 6750
    },
    {
      "entropy": 5.959892606735229,
      "epoch": 0.5965206640762981,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004970877078293483,
      "loss": 5.7856,
      "mean_token_accuracy": 0.16963758170604706,
      "num_tokens": 12585130.0,
      "step": 6755
    },
    {
      "entropy": 6.123706960678101,
      "epoch": 0.5969622041681385,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970826552660111,
      "loss": 5.932,
      "mean_token_accuracy": 0.15625131577253343,
      "num_tokens": 12595598.0,
      "step": 6760
    },
    {
      "entropy": 6.105381727218628,
      "epoch": 0.5974037442599788,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970775983522009,
      "loss": 5.7934,
      "mean_token_accuracy": 0.1726885825395584,
      "num_tokens": 12604905.0,
      "step": 6765
    },
    {
      "entropy": 5.922390365600586,
      "epoch": 0.5978452843518192,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970725370880168,
      "loss": 5.8153,
      "mean_token_accuracy": 0.1768861636519432,
      "num_tokens": 12614683.0,
      "step": 6770
    },
    {
      "entropy": 5.975131320953369,
      "epoch": 0.5982868244436594,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970674714735581,
      "loss": 5.7994,
      "mean_token_accuracy": 0.17253219336271286,
      "num_tokens": 12625199.0,
      "step": 6775
    },
    {
      "entropy": 6.0342090129852295,
      "epoch": 0.5987283645354998,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970624015089239,
      "loss": 5.7283,
      "mean_token_accuracy": 0.1800515815615654,
      "num_tokens": 12634963.0,
      "step": 6780
    },
    {
      "entropy": 5.98271541595459,
      "epoch": 0.5991699046273402,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970573271942135,
      "loss": 5.7174,
      "mean_token_accuracy": 0.18481351882219316,
      "num_tokens": 12643246.0,
      "step": 6785
    },
    {
      "entropy": 5.9873486995697025,
      "epoch": 0.5996114447191805,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970522485295265,
      "loss": 5.8846,
      "mean_token_accuracy": 0.17143715620040895,
      "num_tokens": 12652626.0,
      "step": 6790
    },
    {
      "entropy": 6.0505670547485355,
      "epoch": 0.6000529848110209,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970471655149621,
      "loss": 5.7584,
      "mean_token_accuracy": 0.17371048033237457,
      "num_tokens": 12661337.0,
      "step": 6795
    },
    {
      "entropy": 5.969434070587158,
      "epoch": 0.6004945249028611,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049704207815062,
      "loss": 5.6447,
      "mean_token_accuracy": 0.18933165967464446,
      "num_tokens": 12669716.0,
      "step": 6800
    },
    {
      "entropy": 5.894996881484985,
      "epoch": 0.6009360649947015,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970369864366,
      "loss": 5.7454,
      "mean_token_accuracy": 0.17902431935071944,
      "num_tokens": 12678370.0,
      "step": 6805
    },
    {
      "entropy": 6.023124074935913,
      "epoch": 0.6013776050865418,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970318903730017,
      "loss": 5.7977,
      "mean_token_accuracy": 0.17490170747041703,
      "num_tokens": 12686557.0,
      "step": 6810
    },
    {
      "entropy": 6.02273998260498,
      "epoch": 0.6018191451783822,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970267899599249,
      "loss": 5.6715,
      "mean_token_accuracy": 0.18117174208164216,
      "num_tokens": 12695879.0,
      "step": 6815
    },
    {
      "entropy": 5.934228420257568,
      "epoch": 0.6022606852702226,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970216851974695,
      "loss": 5.6846,
      "mean_token_accuracy": 0.18159392327070237,
      "num_tokens": 12705215.0,
      "step": 6820
    },
    {
      "entropy": 5.983548927307129,
      "epoch": 0.6027022253620629,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970165760857356,
      "loss": 5.8045,
      "mean_token_accuracy": 0.1696965888142586,
      "num_tokens": 12713761.0,
      "step": 6825
    },
    {
      "entropy": 5.942411613464356,
      "epoch": 0.6031437654539032,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970114626248232,
      "loss": 5.6673,
      "mean_token_accuracy": 0.19328989535570146,
      "num_tokens": 12723084.0,
      "step": 6830
    },
    {
      "entropy": 6.048355293273926,
      "epoch": 0.6035853055457435,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004970063448148326,
      "loss": 5.7869,
      "mean_token_accuracy": 0.1724540114402771,
      "num_tokens": 12733070.0,
      "step": 6835
    },
    {
      "entropy": 6.046758222579956,
      "epoch": 0.6040268456375839,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970012226558639,
      "loss": 5.8316,
      "mean_token_accuracy": 0.17009074687957765,
      "num_tokens": 12742967.0,
      "step": 6840
    },
    {
      "entropy": 5.960926961898804,
      "epoch": 0.6044683857294242,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969960961480176,
      "loss": 5.744,
      "mean_token_accuracy": 0.17920432835817338,
      "num_tokens": 12751819.0,
      "step": 6845
    },
    {
      "entropy": 5.958087015151977,
      "epoch": 0.6049099258212646,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004969909652913939,
      "loss": 5.8026,
      "mean_token_accuracy": 0.1734309136867523,
      "num_tokens": 12761718.0,
      "step": 6850
    },
    {
      "entropy": 6.098582077026367,
      "epoch": 0.605351465913105,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969858300860935,
      "loss": 5.8166,
      "mean_token_accuracy": 0.16910268366336823,
      "num_tokens": 12772198.0,
      "step": 6855
    },
    {
      "entropy": 5.9579185962677,
      "epoch": 0.6057930060049452,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004969806905322169,
      "loss": 5.7424,
      "mean_token_accuracy": 0.17886629849672317,
      "num_tokens": 12781718.0,
      "step": 6860
    },
    {
      "entropy": 5.97867488861084,
      "epoch": 0.6062345460967856,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004969755466298649,
      "loss": 5.7994,
      "mean_token_accuracy": 0.17108816057443618,
      "num_tokens": 12790934.0,
      "step": 6865
    },
    {
      "entropy": 6.070623064041138,
      "epoch": 0.6066760861886259,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004969703983791382,
      "loss": 5.8827,
      "mean_token_accuracy": 0.16856029033660888,
      "num_tokens": 12801355.0,
      "step": 6870
    },
    {
      "entropy": 6.110792970657348,
      "epoch": 0.6071176262804663,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004969652457801376,
      "loss": 5.8295,
      "mean_token_accuracy": 0.17313438951969146,
      "num_tokens": 12810811.0,
      "step": 6875
    },
    {
      "entropy": 6.027991485595703,
      "epoch": 0.6075591663723066,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969600888329641,
      "loss": 5.7186,
      "mean_token_accuracy": 0.18415967971086503,
      "num_tokens": 12820608.0,
      "step": 6880
    },
    {
      "entropy": 5.993354463577271,
      "epoch": 0.608000706464147,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004969549275377186,
      "loss": 5.8511,
      "mean_token_accuracy": 0.172364541888237,
      "num_tokens": 12829481.0,
      "step": 6885
    },
    {
      "entropy": 6.024882078170776,
      "epoch": 0.6084422465559873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969497618945024,
      "loss": 5.751,
      "mean_token_accuracy": 0.1737458050251007,
      "num_tokens": 12838914.0,
      "step": 6890
    },
    {
      "entropy": 5.947413444519043,
      "epoch": 0.6088837866478276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969445919034165,
      "loss": 5.6441,
      "mean_token_accuracy": 0.17708998471498488,
      "num_tokens": 12847959.0,
      "step": 6895
    },
    {
      "entropy": 5.989596509933472,
      "epoch": 0.609325326739668,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004969394175645623,
      "loss": 5.7481,
      "mean_token_accuracy": 0.1704538270831108,
      "num_tokens": 12856815.0,
      "step": 6900
    },
    {
      "entropy": 6.024069690704346,
      "epoch": 0.6097668668315083,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969342388780411,
      "loss": 5.788,
      "mean_token_accuracy": 0.1779672086238861,
      "num_tokens": 12866069.0,
      "step": 6905
    },
    {
      "entropy": 5.90646185874939,
      "epoch": 0.6102084069233487,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969290558439546,
      "loss": 5.6222,
      "mean_token_accuracy": 0.18068907409906387,
      "num_tokens": 12874965.0,
      "step": 6910
    },
    {
      "entropy": 5.860468769073487,
      "epoch": 0.6106499470151889,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969238684624038,
      "loss": 5.7142,
      "mean_token_accuracy": 0.1848875477910042,
      "num_tokens": 12883888.0,
      "step": 6915
    },
    {
      "entropy": 5.998827028274536,
      "epoch": 0.6110914871070293,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004969186767334908,
      "loss": 5.796,
      "mean_token_accuracy": 0.17831838577985765,
      "num_tokens": 12893125.0,
      "step": 6920
    },
    {
      "entropy": 6.071118593215942,
      "epoch": 0.6115330271988697,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969134806573172,
      "loss": 5.7661,
      "mean_token_accuracy": 0.17254863828420638,
      "num_tokens": 12902509.0,
      "step": 6925
    },
    {
      "entropy": 6.009656524658203,
      "epoch": 0.61197456729071,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969082802339846,
      "loss": 5.8748,
      "mean_token_accuracy": 0.16720397025346756,
      "num_tokens": 12911624.0,
      "step": 6930
    },
    {
      "entropy": 5.980963039398193,
      "epoch": 0.6124161073825504,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496903075463595,
      "loss": 5.7296,
      "mean_token_accuracy": 0.1769852325320244,
      "num_tokens": 12921022.0,
      "step": 6935
    },
    {
      "entropy": 6.0653139591217045,
      "epoch": 0.6128576474743906,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968978663462504,
      "loss": 5.8553,
      "mean_token_accuracy": 0.1665762558579445,
      "num_tokens": 12930931.0,
      "step": 6940
    },
    {
      "entropy": 5.965591192245483,
      "epoch": 0.613299187566231,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004968926528820529,
      "loss": 5.7101,
      "mean_token_accuracy": 0.18090604543685912,
      "num_tokens": 12939888.0,
      "step": 6945
    },
    {
      "entropy": 5.9705089092254635,
      "epoch": 0.6137407276580713,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968874350711044,
      "loss": 5.769,
      "mean_token_accuracy": 0.17528013586997987,
      "num_tokens": 12948243.0,
      "step": 6950
    },
    {
      "entropy": 6.028779077529907,
      "epoch": 0.6141822677499117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968822129135072,
      "loss": 5.8194,
      "mean_token_accuracy": 0.17186571657657623,
      "num_tokens": 12957612.0,
      "step": 6955
    },
    {
      "entropy": 6.033588314056397,
      "epoch": 0.6146238078417521,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968769864093637,
      "loss": 5.7815,
      "mean_token_accuracy": 0.17663713544607162,
      "num_tokens": 12966926.0,
      "step": 6960
    },
    {
      "entropy": 5.999309158325195,
      "epoch": 0.6150653479335924,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004968717555587762,
      "loss": 5.8454,
      "mean_token_accuracy": 0.1690806269645691,
      "num_tokens": 12976630.0,
      "step": 6965
    },
    {
      "entropy": 5.995844173431396,
      "epoch": 0.6155068880254327,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968665203618472,
      "loss": 5.7905,
      "mean_token_accuracy": 0.17869388163089753,
      "num_tokens": 12986458.0,
      "step": 6970
    },
    {
      "entropy": 6.046888399124145,
      "epoch": 0.615948428117273,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968612808186793,
      "loss": 5.8812,
      "mean_token_accuracy": 0.16704486906528473,
      "num_tokens": 12995564.0,
      "step": 6975
    },
    {
      "entropy": 6.09203953742981,
      "epoch": 0.6163899682091134,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004968560369293751,
      "loss": 5.8564,
      "mean_token_accuracy": 0.16267934888601304,
      "num_tokens": 13004920.0,
      "step": 6980
    },
    {
      "entropy": 5.961414813995361,
      "epoch": 0.6168315083009537,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004968507886940372,
      "loss": 5.7313,
      "mean_token_accuracy": 0.18028314113616944,
      "num_tokens": 13013967.0,
      "step": 6985
    },
    {
      "entropy": 5.927100419998169,
      "epoch": 0.6172730483927941,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968455361127686,
      "loss": 5.7644,
      "mean_token_accuracy": 0.18007203787565232,
      "num_tokens": 13023395.0,
      "step": 6990
    },
    {
      "entropy": 6.0053568363189695,
      "epoch": 0.6177145884846345,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968402791856721,
      "loss": 5.7687,
      "mean_token_accuracy": 0.1733642503619194,
      "num_tokens": 13032792.0,
      "step": 6995
    },
    {
      "entropy": 6.0467626571655275,
      "epoch": 0.6181561285764747,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968350179128507,
      "loss": 5.7777,
      "mean_token_accuracy": 0.1806527256965637,
      "num_tokens": 13043425.0,
      "step": 7000
    },
    {
      "entropy": 6.011540651321411,
      "epoch": 0.6185976686683151,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004968297522944075,
      "loss": 5.8461,
      "mean_token_accuracy": 0.1733243554830551,
      "num_tokens": 13052472.0,
      "step": 7005
    },
    {
      "entropy": 6.023688936233521,
      "epoch": 0.6190392087601554,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004968244823304455,
      "loss": 5.7682,
      "mean_token_accuracy": 0.17574435621500015,
      "num_tokens": 13062057.0,
      "step": 7010
    },
    {
      "entropy": 6.0068443775177,
      "epoch": 0.6194807488519958,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968192080210683,
      "loss": 5.6998,
      "mean_token_accuracy": 0.1804433837532997,
      "num_tokens": 13070751.0,
      "step": 7015
    },
    {
      "entropy": 5.866034269332886,
      "epoch": 0.619922288943836,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968139293663787,
      "loss": 5.6124,
      "mean_token_accuracy": 0.19116149544715882,
      "num_tokens": 13080794.0,
      "step": 7020
    },
    {
      "entropy": 5.948003339767456,
      "epoch": 0.6203638290356764,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968086463664806,
      "loss": 5.7106,
      "mean_token_accuracy": 0.1824044868350029,
      "num_tokens": 13090120.0,
      "step": 7025
    },
    {
      "entropy": 5.999112796783447,
      "epoch": 0.6208053691275168,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496803359021477,
      "loss": 5.7563,
      "mean_token_accuracy": 0.17223722636699676,
      "num_tokens": 13099295.0,
      "step": 7030
    },
    {
      "entropy": 5.930928897857666,
      "epoch": 0.6212469092193571,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967980673314719,
      "loss": 5.7683,
      "mean_token_accuracy": 0.18261488974094392,
      "num_tokens": 13107509.0,
      "step": 7035
    },
    {
      "entropy": 5.902431058883667,
      "epoch": 0.6216884493111975,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967927712965688,
      "loss": 5.7512,
      "mean_token_accuracy": 0.17700175195932388,
      "num_tokens": 13116991.0,
      "step": 7040
    },
    {
      "entropy": 6.0278968334198,
      "epoch": 0.6221299894030378,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967874709168714,
      "loss": 5.8274,
      "mean_token_accuracy": 0.1767498642206192,
      "num_tokens": 13126362.0,
      "step": 7045
    },
    {
      "entropy": 6.070158243179321,
      "epoch": 0.6225715294948782,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967821661924837,
      "loss": 5.7047,
      "mean_token_accuracy": 0.18593760579824448,
      "num_tokens": 13135938.0,
      "step": 7050
    },
    {
      "entropy": 5.993052244186401,
      "epoch": 0.6230130695867184,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967768571235094,
      "loss": 5.7948,
      "mean_token_accuracy": 0.17593481093645097,
      "num_tokens": 13145516.0,
      "step": 7055
    },
    {
      "entropy": 5.916328859329224,
      "epoch": 0.6234546096785588,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967715437100526,
      "loss": 5.7078,
      "mean_token_accuracy": 0.17639447450637818,
      "num_tokens": 13155114.0,
      "step": 7060
    },
    {
      "entropy": 6.005270290374756,
      "epoch": 0.6238961497703992,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967662259522174,
      "loss": 5.6839,
      "mean_token_accuracy": 0.18816752284765242,
      "num_tokens": 13164759.0,
      "step": 7065
    },
    {
      "entropy": 5.956115245819092,
      "epoch": 0.6243376898622395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000496760903850108,
      "loss": 5.7326,
      "mean_token_accuracy": 0.1707065910100937,
      "num_tokens": 13173752.0,
      "step": 7070
    },
    {
      "entropy": 5.97244873046875,
      "epoch": 0.6247792299540799,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967555774038286,
      "loss": 5.7648,
      "mean_token_accuracy": 0.172952039539814,
      "num_tokens": 13183004.0,
      "step": 7075
    },
    {
      "entropy": 5.991962003707886,
      "epoch": 0.6252207700459201,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967502466134836,
      "loss": 5.7862,
      "mean_token_accuracy": 0.17271161526441575,
      "num_tokens": 13192201.0,
      "step": 7080
    },
    {
      "entropy": 5.991434097290039,
      "epoch": 0.6256623101377605,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004967449114791773,
      "loss": 5.809,
      "mean_token_accuracy": 0.17219959050416947,
      "num_tokens": 13201419.0,
      "step": 7085
    },
    {
      "entropy": 5.9468615531921385,
      "epoch": 0.6261038502296008,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967395720010143,
      "loss": 5.5955,
      "mean_token_accuracy": 0.19906508028507233,
      "num_tokens": 13210073.0,
      "step": 7090
    },
    {
      "entropy": 5.982910537719727,
      "epoch": 0.6265453903214412,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004967342281790992,
      "loss": 5.8127,
      "mean_token_accuracy": 0.17153889387845994,
      "num_tokens": 13219768.0,
      "step": 7095
    },
    {
      "entropy": 6.003438901901245,
      "epoch": 0.6269869304132816,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004967288800135368,
      "loss": 5.6807,
      "mean_token_accuracy": 0.17888256758451462,
      "num_tokens": 13229439.0,
      "step": 7100
    },
    {
      "entropy": 5.9994972229003904,
      "epoch": 0.6274284705051218,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967235275044315,
      "loss": 5.6548,
      "mean_token_accuracy": 0.18393336832523347,
      "num_tokens": 13238067.0,
      "step": 7105
    },
    {
      "entropy": 5.977489805221557,
      "epoch": 0.6278700105969622,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967181706518886,
      "loss": 5.753,
      "mean_token_accuracy": 0.17324308604001998,
      "num_tokens": 13246758.0,
      "step": 7110
    },
    {
      "entropy": 5.953913974761963,
      "epoch": 0.6283115506888025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004967128094560128,
      "loss": 5.7616,
      "mean_token_accuracy": 0.18290273547172547,
      "num_tokens": 13255510.0,
      "step": 7115
    },
    {
      "entropy": 5.957019090652466,
      "epoch": 0.6287530907806429,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967074439169092,
      "loss": 5.7223,
      "mean_token_accuracy": 0.17530591785907745,
      "num_tokens": 13265629.0,
      "step": 7120
    },
    {
      "entropy": 5.9987959384918215,
      "epoch": 0.6291946308724832,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004967020740346827,
      "loss": 5.7624,
      "mean_token_accuracy": 0.1819077268242836,
      "num_tokens": 13275871.0,
      "step": 7125
    },
    {
      "entropy": 6.087621402740479,
      "epoch": 0.6296361709643236,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966966998094387,
      "loss": 5.8294,
      "mean_token_accuracy": 0.16260403543710708,
      "num_tokens": 13285134.0,
      "step": 7130
    },
    {
      "entropy": 5.9442079067230225,
      "epoch": 0.630077711056164,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966913212412825,
      "loss": 5.7328,
      "mean_token_accuracy": 0.182009693980217,
      "num_tokens": 13294127.0,
      "step": 7135
    },
    {
      "entropy": 6.0073131084442135,
      "epoch": 0.6305192511480042,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966859383303194,
      "loss": 5.8638,
      "mean_token_accuracy": 0.16730156242847444,
      "num_tokens": 13303847.0,
      "step": 7140
    },
    {
      "entropy": 6.057704877853394,
      "epoch": 0.6309607912398446,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966805510766549,
      "loss": 5.8814,
      "mean_token_accuracy": 0.17394553869962692,
      "num_tokens": 13313267.0,
      "step": 7145
    },
    {
      "entropy": 6.015100336074829,
      "epoch": 0.6314023313316849,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966751594803944,
      "loss": 5.7636,
      "mean_token_accuracy": 0.18172619193792344,
      "num_tokens": 13321920.0,
      "step": 7150
    },
    {
      "entropy": 5.979697418212891,
      "epoch": 0.6318438714235253,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004966697635416437,
      "loss": 5.7735,
      "mean_token_accuracy": 0.17064696699380874,
      "num_tokens": 13331953.0,
      "step": 7155
    },
    {
      "entropy": 6.0059898853302,
      "epoch": 0.6322854115153655,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966643632605082,
      "loss": 5.6716,
      "mean_token_accuracy": 0.18357742428779603,
      "num_tokens": 13340353.0,
      "step": 7160
    },
    {
      "entropy": 5.990286493301392,
      "epoch": 0.6327269516072059,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000496658958637094,
      "loss": 5.7902,
      "mean_token_accuracy": 0.16551219522953034,
      "num_tokens": 13350324.0,
      "step": 7165
    },
    {
      "entropy": 5.942095947265625,
      "epoch": 0.6331684916990463,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496653549671507,
      "loss": 5.6891,
      "mean_token_accuracy": 0.1809837579727173,
      "num_tokens": 13359233.0,
      "step": 7170
    },
    {
      "entropy": 5.949536657333374,
      "epoch": 0.6336100317908866,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004966481363638529,
      "loss": 5.731,
      "mean_token_accuracy": 0.1728736624121666,
      "num_tokens": 13368675.0,
      "step": 7175
    },
    {
      "entropy": 6.0458451271057125,
      "epoch": 0.634051571882727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966427187142378,
      "loss": 5.7287,
      "mean_token_accuracy": 0.17600933462381363,
      "num_tokens": 13377467.0,
      "step": 7180
    },
    {
      "entropy": 6.050792980194092,
      "epoch": 0.6344931119745673,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004966372967227681,
      "loss": 5.9016,
      "mean_token_accuracy": 0.16676654815673828,
      "num_tokens": 13388194.0,
      "step": 7185
    },
    {
      "entropy": 5.948937702178955,
      "epoch": 0.6349346520664076,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966318703895496,
      "loss": 5.6191,
      "mean_token_accuracy": 0.18056136965751649,
      "num_tokens": 13397115.0,
      "step": 7190
    },
    {
      "entropy": 5.9974937915802,
      "epoch": 0.6353761921582479,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966264397146888,
      "loss": 5.7093,
      "mean_token_accuracy": 0.17761638760566711,
      "num_tokens": 13405860.0,
      "step": 7195
    },
    {
      "entropy": 5.938602161407471,
      "epoch": 0.6358177322500883,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966210046982921,
      "loss": 5.7088,
      "mean_token_accuracy": 0.18449603021144867,
      "num_tokens": 13415489.0,
      "step": 7200
    },
    {
      "entropy": 5.990828561782837,
      "epoch": 0.6362592723419287,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966155653404661,
      "loss": 5.7638,
      "mean_token_accuracy": 0.1800005167722702,
      "num_tokens": 13424780.0,
      "step": 7205
    },
    {
      "entropy": 5.9898899555206295,
      "epoch": 0.636700812433769,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966101216413172,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17806257009506227,
      "num_tokens": 13433835.0,
      "step": 7210
    },
    {
      "entropy": 5.92235426902771,
      "epoch": 0.6371423525256094,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004966046736009519,
      "loss": 5.673,
      "mean_token_accuracy": 0.18577371686697006,
      "num_tokens": 13443054.0,
      "step": 7215
    },
    {
      "entropy": 5.919622755050659,
      "epoch": 0.6375838926174496,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965992212194772,
      "loss": 5.68,
      "mean_token_accuracy": 0.18048593103885652,
      "num_tokens": 13452148.0,
      "step": 7220
    },
    {
      "entropy": 5.988049745559692,
      "epoch": 0.63802543270929,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965937644969998,
      "loss": 5.6613,
      "mean_token_accuracy": 0.18048649728298188,
      "num_tokens": 13461015.0,
      "step": 7225
    },
    {
      "entropy": 5.958803796768189,
      "epoch": 0.6384669728011303,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965883034336266,
      "loss": 5.7926,
      "mean_token_accuracy": 0.17370286136865615,
      "num_tokens": 13470708.0,
      "step": 7230
    },
    {
      "entropy": 5.929748821258545,
      "epoch": 0.6389085128929707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965828380294645,
      "loss": 5.647,
      "mean_token_accuracy": 0.18508279025554658,
      "num_tokens": 13479469.0,
      "step": 7235
    },
    {
      "entropy": 6.008753395080566,
      "epoch": 0.6393500529848111,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965773682846207,
      "loss": 5.75,
      "mean_token_accuracy": 0.1739913746714592,
      "num_tokens": 13488575.0,
      "step": 7240
    },
    {
      "entropy": 5.964322328567505,
      "epoch": 0.6397915930766513,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965718941992021,
      "loss": 5.7357,
      "mean_token_accuracy": 0.17797115296125413,
      "num_tokens": 13497539.0,
      "step": 7245
    },
    {
      "entropy": 6.023652648925781,
      "epoch": 0.6402331331684917,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965664157733162,
      "loss": 5.7795,
      "mean_token_accuracy": 0.18303052335977554,
      "num_tokens": 13506159.0,
      "step": 7250
    },
    {
      "entropy": 5.9235453605651855,
      "epoch": 0.640674673260332,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965609330070703,
      "loss": 5.7725,
      "mean_token_accuracy": 0.17385415583848954,
      "num_tokens": 13516415.0,
      "step": 7255
    },
    {
      "entropy": 6.039669322967529,
      "epoch": 0.6411162133521724,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965554459005717,
      "loss": 5.8067,
      "mean_token_accuracy": 0.17283896803855897,
      "num_tokens": 13525433.0,
      "step": 7260
    },
    {
      "entropy": 6.055001831054687,
      "epoch": 0.6415577534440127,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004965499544539278,
      "loss": 5.7811,
      "mean_token_accuracy": 0.1746595412492752,
      "num_tokens": 13535173.0,
      "step": 7265
    },
    {
      "entropy": 5.923309087753296,
      "epoch": 0.641999293535853,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965444586672463,
      "loss": 5.7171,
      "mean_token_accuracy": 0.1860516145825386,
      "num_tokens": 13544280.0,
      "step": 7270
    },
    {
      "entropy": 6.037317276000977,
      "epoch": 0.6424408336276934,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496538958540635,
      "loss": 5.6718,
      "mean_token_accuracy": 0.18603464663028718,
      "num_tokens": 13553358.0,
      "step": 7275
    },
    {
      "entropy": 6.019790840148926,
      "epoch": 0.6428823737195337,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004965334540742015,
      "loss": 5.8651,
      "mean_token_accuracy": 0.169008007645607,
      "num_tokens": 13562904.0,
      "step": 7280
    },
    {
      "entropy": 5.930609560012817,
      "epoch": 0.6433239138113741,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004965279452680535,
      "loss": 5.6597,
      "mean_token_accuracy": 0.18698660880327225,
      "num_tokens": 13572745.0,
      "step": 7285
    },
    {
      "entropy": 5.989558744430542,
      "epoch": 0.6437654539032144,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004965224321222991,
      "loss": 5.7325,
      "mean_token_accuracy": 0.18167387545108796,
      "num_tokens": 13581515.0,
      "step": 7290
    },
    {
      "entropy": 5.987615871429443,
      "epoch": 0.6442069939950548,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965169146370462,
      "loss": 5.8105,
      "mean_token_accuracy": 0.1702969953417778,
      "num_tokens": 13591052.0,
      "step": 7295
    },
    {
      "entropy": 6.006259393692017,
      "epoch": 0.644648534086895,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965113928124031,
      "loss": 5.6803,
      "mean_token_accuracy": 0.17826583236455917,
      "num_tokens": 13599790.0,
      "step": 7300
    },
    {
      "entropy": 5.92980728149414,
      "epoch": 0.6450900741787354,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965058666484775,
      "loss": 5.6924,
      "mean_token_accuracy": 0.17433585077524186,
      "num_tokens": 13609692.0,
      "step": 7305
    },
    {
      "entropy": 5.919853973388672,
      "epoch": 0.6455316142705758,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965003361453782,
      "loss": 5.6959,
      "mean_token_accuracy": 0.17826921194791795,
      "num_tokens": 13617857.0,
      "step": 7310
    },
    {
      "entropy": 5.999902391433716,
      "epoch": 0.6459731543624161,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964948013032132,
      "loss": 5.7661,
      "mean_token_accuracy": 0.18131729662418367,
      "num_tokens": 13626954.0,
      "step": 7315
    },
    {
      "entropy": 6.036501264572143,
      "epoch": 0.6464146944542565,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496489262122091,
      "loss": 5.8375,
      "mean_token_accuracy": 0.17557971626520158,
      "num_tokens": 13637075.0,
      "step": 7320
    },
    {
      "entropy": 6.031018686294556,
      "epoch": 0.6468562345460968,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964837186021201,
      "loss": 5.7166,
      "mean_token_accuracy": 0.183514204621315,
      "num_tokens": 13646364.0,
      "step": 7325
    },
    {
      "entropy": 5.904260063171387,
      "epoch": 0.6472977746379371,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004964781707434093,
      "loss": 5.6658,
      "mean_token_accuracy": 0.18607676029205322,
      "num_tokens": 13655312.0,
      "step": 7330
    },
    {
      "entropy": 5.9440549373626705,
      "epoch": 0.6477393147297774,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964726185460669,
      "loss": 5.7509,
      "mean_token_accuracy": 0.17000489085912704,
      "num_tokens": 13664049.0,
      "step": 7335
    },
    {
      "entropy": 5.979244232177734,
      "epoch": 0.6481808548216178,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004964670620102019,
      "loss": 5.6325,
      "mean_token_accuracy": 0.19036355167627333,
      "num_tokens": 13673028.0,
      "step": 7340
    },
    {
      "entropy": 5.998539876937866,
      "epoch": 0.6486223949134582,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964615011359231,
      "loss": 5.7026,
      "mean_token_accuracy": 0.1783529758453369,
      "num_tokens": 13682258.0,
      "step": 7345
    },
    {
      "entropy": 5.9948320388793945,
      "epoch": 0.6490639350052985,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004964559359233395,
      "loss": 5.732,
      "mean_token_accuracy": 0.17761077731847763,
      "num_tokens": 13691237.0,
      "step": 7350
    },
    {
      "entropy": 5.970560073852539,
      "epoch": 0.6495054750971389,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004964503663725601,
      "loss": 5.7183,
      "mean_token_accuracy": 0.18295522034168243,
      "num_tokens": 13701385.0,
      "step": 7355
    },
    {
      "entropy": 5.9920416355133055,
      "epoch": 0.6499470151889791,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964447924836938,
      "loss": 5.7305,
      "mean_token_accuracy": 0.17568194717168809,
      "num_tokens": 13711187.0,
      "step": 7360
    },
    {
      "entropy": 5.911662149429321,
      "epoch": 0.6503885552808195,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004964392142568501,
      "loss": 5.7021,
      "mean_token_accuracy": 0.1853645294904709,
      "num_tokens": 13720778.0,
      "step": 7365
    },
    {
      "entropy": 5.9825903415679935,
      "epoch": 0.6508300953726598,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004964336316921381,
      "loss": 5.7987,
      "mean_token_accuracy": 0.17075325101613997,
      "num_tokens": 13730172.0,
      "step": 7370
    },
    {
      "entropy": 5.909342241287232,
      "epoch": 0.6512716354645002,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964280447896672,
      "loss": 5.6056,
      "mean_token_accuracy": 0.18496739566326142,
      "num_tokens": 13739191.0,
      "step": 7375
    },
    {
      "entropy": 5.8662872314453125,
      "epoch": 0.6517131755563406,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964224535495468,
      "loss": 5.7015,
      "mean_token_accuracy": 0.18262420296669007,
      "num_tokens": 13748168.0,
      "step": 7380
    },
    {
      "entropy": 6.014463663101196,
      "epoch": 0.6521547156481808,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004964168579718864,
      "loss": 5.7744,
      "mean_token_accuracy": 0.17843029499053956,
      "num_tokens": 13758334.0,
      "step": 7385
    },
    {
      "entropy": 5.984301328659058,
      "epoch": 0.6525962557400212,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004964112580567959,
      "loss": 5.6231,
      "mean_token_accuracy": 0.17938115149736406,
      "num_tokens": 13767768.0,
      "step": 7390
    },
    {
      "entropy": 5.972589731216431,
      "epoch": 0.6530377958318615,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004964056538043845,
      "loss": 5.7117,
      "mean_token_accuracy": 0.18241407722234726,
      "num_tokens": 13776935.0,
      "step": 7395
    },
    {
      "entropy": 5.9949897766113285,
      "epoch": 0.6534793359237019,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964000452147624,
      "loss": 5.7772,
      "mean_token_accuracy": 0.17291238605976106,
      "num_tokens": 13786539.0,
      "step": 7400
    },
    {
      "entropy": 5.936427068710327,
      "epoch": 0.6539208760155422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963944322880394,
      "loss": 5.6811,
      "mean_token_accuracy": 0.1794592872262001,
      "num_tokens": 13794920.0,
      "step": 7405
    },
    {
      "entropy": 6.007709264755249,
      "epoch": 0.6543624161073825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004963888150243253,
      "loss": 5.8283,
      "mean_token_accuracy": 0.17710065543651582,
      "num_tokens": 13804577.0,
      "step": 7410
    },
    {
      "entropy": 6.0295015335083,
      "epoch": 0.6548039561992229,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963831934237302,
      "loss": 5.7007,
      "mean_token_accuracy": 0.1782134383916855,
      "num_tokens": 13813485.0,
      "step": 7415
    },
    {
      "entropy": 5.969497537612915,
      "epoch": 0.6552454962910632,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004963775674863643,
      "loss": 5.6537,
      "mean_token_accuracy": 0.18231142908334733,
      "num_tokens": 13821935.0,
      "step": 7420
    },
    {
      "entropy": 5.904452323913574,
      "epoch": 0.6556870363829036,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004963719372123378,
      "loss": 5.742,
      "mean_token_accuracy": 0.17926856577396394,
      "num_tokens": 13831471.0,
      "step": 7425
    },
    {
      "entropy": 5.9243052959442135,
      "epoch": 0.6561285764747439,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963663026017609,
      "loss": 5.6757,
      "mean_token_accuracy": 0.1833261579275131,
      "num_tokens": 13841105.0,
      "step": 7430
    },
    {
      "entropy": 6.02013750076294,
      "epoch": 0.6565701165665843,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963606636547441,
      "loss": 5.7312,
      "mean_token_accuracy": 0.17202972173690795,
      "num_tokens": 13850303.0,
      "step": 7435
    },
    {
      "entropy": 5.881796026229859,
      "epoch": 0.6570116566584245,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963550203713976,
      "loss": 5.6301,
      "mean_token_accuracy": 0.19145385921001434,
      "num_tokens": 13860684.0,
      "step": 7440
    },
    {
      "entropy": 5.995109128952026,
      "epoch": 0.6574531967502649,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963493727518323,
      "loss": 5.7879,
      "mean_token_accuracy": 0.176690474152565,
      "num_tokens": 13870258.0,
      "step": 7445
    },
    {
      "entropy": 6.027615261077881,
      "epoch": 0.6578947368421053,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004963437207961585,
      "loss": 5.7019,
      "mean_token_accuracy": 0.17877504825592042,
      "num_tokens": 13878956.0,
      "step": 7450
    },
    {
      "entropy": 5.952871465682984,
      "epoch": 0.6583362769339456,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963380645044874,
      "loss": 5.7398,
      "mean_token_accuracy": 0.17886650264263154,
      "num_tokens": 13888614.0,
      "step": 7455
    },
    {
      "entropy": 5.992289400100708,
      "epoch": 0.658777817025786,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963324038769292,
      "loss": 5.7658,
      "mean_token_accuracy": 0.18047383278608323,
      "num_tokens": 13898035.0,
      "step": 7460
    },
    {
      "entropy": 5.973392963409424,
      "epoch": 0.6592193571176262,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004963267389135953,
      "loss": 5.7837,
      "mean_token_accuracy": 0.16953325122594834,
      "num_tokens": 13908107.0,
      "step": 7465
    },
    {
      "entropy": 5.964011144638062,
      "epoch": 0.6596608972094666,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004963210696145964,
      "loss": 5.6493,
      "mean_token_accuracy": 0.18856916427612305,
      "num_tokens": 13917149.0,
      "step": 7470
    },
    {
      "entropy": 6.0005358219146725,
      "epoch": 0.6601024373013069,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004963153959800436,
      "loss": 5.7416,
      "mean_token_accuracy": 0.17932893037796022,
      "num_tokens": 13926465.0,
      "step": 7475
    },
    {
      "entropy": 5.96647367477417,
      "epoch": 0.6605439773931473,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000496309718010048,
      "loss": 5.7748,
      "mean_token_accuracy": 0.17536671906709672,
      "num_tokens": 13936109.0,
      "step": 7480
    },
    {
      "entropy": 5.9665961265563965,
      "epoch": 0.6609855174849877,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963040357047209,
      "loss": 5.7129,
      "mean_token_accuracy": 0.17488997280597687,
      "num_tokens": 13946227.0,
      "step": 7485
    },
    {
      "entropy": 6.044988441467285,
      "epoch": 0.661427057576828,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004962983490641737,
      "loss": 5.8221,
      "mean_token_accuracy": 0.17853489369153977,
      "num_tokens": 13955908.0,
      "step": 7490
    },
    {
      "entropy": 6.025640058517456,
      "epoch": 0.6618685976686683,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004962926580885177,
      "loss": 5.725,
      "mean_token_accuracy": 0.17798093408346177,
      "num_tokens": 13964736.0,
      "step": 7495
    },
    {
      "entropy": 6.042965936660766,
      "epoch": 0.6623101377605086,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004962869627778642,
      "loss": 5.7862,
      "mean_token_accuracy": 0.16908291578292847,
      "num_tokens": 13974980.0,
      "step": 7500
    },
    {
      "entropy": 6.118487596511841,
      "epoch": 0.662751677852349,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962812631323252,
      "loss": 5.8703,
      "mean_token_accuracy": 0.17226627022027968,
      "num_tokens": 13984473.0,
      "step": 7505
    },
    {
      "entropy": 5.970974254608154,
      "epoch": 0.6631932179441893,
      "grad_norm": 0.84375,
      "learning_rate": 0.000496275559152012,
      "loss": 5.6888,
      "mean_token_accuracy": 0.18050990998744965,
      "num_tokens": 13994151.0,
      "step": 7510
    },
    {
      "entropy": 6.025629663467408,
      "epoch": 0.6636347580360297,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004962698508370365,
      "loss": 5.7853,
      "mean_token_accuracy": 0.17294417917728425,
      "num_tokens": 14003347.0,
      "step": 7515
    },
    {
      "entropy": 5.935830354690552,
      "epoch": 0.6640762981278701,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004962641381875103,
      "loss": 5.6647,
      "mean_token_accuracy": 0.18774728178977967,
      "num_tokens": 14012817.0,
      "step": 7520
    },
    {
      "entropy": 5.948458099365235,
      "epoch": 0.6645178382197103,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962584212035456,
      "loss": 5.7051,
      "mean_token_accuracy": 0.17370816171169282,
      "num_tokens": 14021915.0,
      "step": 7525
    },
    {
      "entropy": 5.891166257858276,
      "epoch": 0.6649593783115507,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962526998852543,
      "loss": 5.6905,
      "mean_token_accuracy": 0.1777677059173584,
      "num_tokens": 14030848.0,
      "step": 7530
    },
    {
      "entropy": 5.924250316619873,
      "epoch": 0.665400918403391,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962469742327483,
      "loss": 5.7048,
      "mean_token_accuracy": 0.18236329704523085,
      "num_tokens": 14040292.0,
      "step": 7535
    },
    {
      "entropy": 6.023686361312866,
      "epoch": 0.6658424584952314,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049624124424614,
      "loss": 5.8033,
      "mean_token_accuracy": 0.17605432122945786,
      "num_tokens": 14048829.0,
      "step": 7540
    },
    {
      "entropy": 5.928983402252197,
      "epoch": 0.6662839985870717,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962355099255415,
      "loss": 5.6705,
      "mean_token_accuracy": 0.1892841041088104,
      "num_tokens": 14058863.0,
      "step": 7545
    },
    {
      "entropy": 5.927666759490966,
      "epoch": 0.666725538678912,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962297712710653,
      "loss": 5.708,
      "mean_token_accuracy": 0.18500306606292724,
      "num_tokens": 14068184.0,
      "step": 7550
    },
    {
      "entropy": 5.957968711853027,
      "epoch": 0.6671670787707524,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004962240282828236,
      "loss": 5.7257,
      "mean_token_accuracy": 0.176913782954216,
      "num_tokens": 14078626.0,
      "step": 7555
    },
    {
      "entropy": 6.025596714019775,
      "epoch": 0.6676086188625927,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496218280960929,
      "loss": 5.7506,
      "mean_token_accuracy": 0.17131045907735826,
      "num_tokens": 14087615.0,
      "step": 7560
    },
    {
      "entropy": 5.970422744750977,
      "epoch": 0.6680501589544331,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962125293054941,
      "loss": 5.7448,
      "mean_token_accuracy": 0.18081327825784682,
      "num_tokens": 14097769.0,
      "step": 7565
    },
    {
      "entropy": 5.969198274612427,
      "epoch": 0.6684916990462734,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962067733166317,
      "loss": 5.6805,
      "mean_token_accuracy": 0.18027809858322144,
      "num_tokens": 14106761.0,
      "step": 7570
    },
    {
      "entropy": 5.842503023147583,
      "epoch": 0.6689332391381138,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962010129944542,
      "loss": 5.5288,
      "mean_token_accuracy": 0.19351110756397247,
      "num_tokens": 14116333.0,
      "step": 7575
    },
    {
      "entropy": 5.920039796829224,
      "epoch": 0.669374779229954,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961952483390747,
      "loss": 5.6346,
      "mean_token_accuracy": 0.18978767842054367,
      "num_tokens": 14125927.0,
      "step": 7580
    },
    {
      "entropy": 5.9154970169067385,
      "epoch": 0.6698163193217944,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961894793506062,
      "loss": 5.5979,
      "mean_token_accuracy": 0.1923290029168129,
      "num_tokens": 14135069.0,
      "step": 7585
    },
    {
      "entropy": 5.901021480560303,
      "epoch": 0.6702578594136348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961837060291615,
      "loss": 5.779,
      "mean_token_accuracy": 0.17629592418670653,
      "num_tokens": 14144889.0,
      "step": 7590
    },
    {
      "entropy": 6.086011743545532,
      "epoch": 0.6706993995054751,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004961779283748538,
      "loss": 5.8326,
      "mean_token_accuracy": 0.16746109426021577,
      "num_tokens": 14154614.0,
      "step": 7595
    },
    {
      "entropy": 5.975073003768921,
      "epoch": 0.6711409395973155,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004961721463877963,
      "loss": 5.7161,
      "mean_token_accuracy": 0.17576538622379304,
      "num_tokens": 14164311.0,
      "step": 7600
    },
    {
      "entropy": 5.919068002700806,
      "epoch": 0.6715824796891557,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961663600681023,
      "loss": 5.6916,
      "mean_token_accuracy": 0.18112306147813798,
      "num_tokens": 14173422.0,
      "step": 7605
    },
    {
      "entropy": 5.935012435913086,
      "epoch": 0.6720240197809961,
      "grad_norm": 0.90625,
      "learning_rate": 0.000496160569415885,
      "loss": 5.7402,
      "mean_token_accuracy": 0.17910145670175553,
      "num_tokens": 14183237.0,
      "step": 7610
    },
    {
      "entropy": 6.006856441497803,
      "epoch": 0.6724655598728364,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961547744312579,
      "loss": 5.6633,
      "mean_token_accuracy": 0.17394500076770783,
      "num_tokens": 14192899.0,
      "step": 7615
    },
    {
      "entropy": 5.962245082855224,
      "epoch": 0.6729070999646768,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961489751143346,
      "loss": 5.7549,
      "mean_token_accuracy": 0.1785586342215538,
      "num_tokens": 14202451.0,
      "step": 7620
    },
    {
      "entropy": 5.950581884384155,
      "epoch": 0.6733486400565172,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961431714652287,
      "loss": 5.5971,
      "mean_token_accuracy": 0.20044878125190735,
      "num_tokens": 14211105.0,
      "step": 7625
    },
    {
      "entropy": 5.945085430145264,
      "epoch": 0.6737901801483575,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961373634840538,
      "loss": 5.7475,
      "mean_token_accuracy": 0.18027413189411162,
      "num_tokens": 14221650.0,
      "step": 7630
    },
    {
      "entropy": 5.962568044662476,
      "epoch": 0.6742317202401978,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961315511709237,
      "loss": 5.6487,
      "mean_token_accuracy": 0.1884300261735916,
      "num_tokens": 14231104.0,
      "step": 7635
    },
    {
      "entropy": 5.957081937789917,
      "epoch": 0.6746732603320381,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004961257345259524,
      "loss": 5.7484,
      "mean_token_accuracy": 0.17803122699260712,
      "num_tokens": 14240355.0,
      "step": 7640
    },
    {
      "entropy": 5.959647512435913,
      "epoch": 0.6751148004238785,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961199135492535,
      "loss": 5.6713,
      "mean_token_accuracy": 0.18485748916864395,
      "num_tokens": 14250065.0,
      "step": 7645
    },
    {
      "entropy": 5.911368799209595,
      "epoch": 0.6755563405157188,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004961140882409414,
      "loss": 5.6349,
      "mean_token_accuracy": 0.1868532657623291,
      "num_tokens": 14259265.0,
      "step": 7650
    },
    {
      "entropy": 5.999258613586425,
      "epoch": 0.6759978806075592,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049610825860113,
      "loss": 5.7861,
      "mean_token_accuracy": 0.18090734332799913,
      "num_tokens": 14269569.0,
      "step": 7655
    },
    {
      "entropy": 5.97138237953186,
      "epoch": 0.6764394206993996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961024246299336,
      "loss": 5.741,
      "mean_token_accuracy": 0.1716483399271965,
      "num_tokens": 14279263.0,
      "step": 7660
    },
    {
      "entropy": 6.00728850364685,
      "epoch": 0.6768809607912398,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960965863274665,
      "loss": 5.7377,
      "mean_token_accuracy": 0.17848307341337205,
      "num_tokens": 14288310.0,
      "step": 7665
    },
    {
      "entropy": 5.918211936950684,
      "epoch": 0.6773225008830802,
      "grad_norm": 0.90625,
      "learning_rate": 0.000496090743693843,
      "loss": 5.6955,
      "mean_token_accuracy": 0.1824905753135681,
      "num_tokens": 14297726.0,
      "step": 7670
    },
    {
      "entropy": 6.054973173141479,
      "epoch": 0.6777640409749205,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004960848967291775,
      "loss": 5.8622,
      "mean_token_accuracy": 0.17365490794181823,
      "num_tokens": 14307247.0,
      "step": 7675
    },
    {
      "entropy": 6.016622877120971,
      "epoch": 0.6782055810667609,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960790454335846,
      "loss": 5.7316,
      "mean_token_accuracy": 0.17730930894613267,
      "num_tokens": 14316861.0,
      "step": 7680
    },
    {
      "entropy": 5.949781513214111,
      "epoch": 0.6786471211586012,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496073189807179,
      "loss": 5.6716,
      "mean_token_accuracy": 0.18665842115879058,
      "num_tokens": 14325898.0,
      "step": 7685
    },
    {
      "entropy": 6.0222639560699465,
      "epoch": 0.6790886612504415,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960673298500753,
      "loss": 5.7763,
      "mean_token_accuracy": 0.17348007261753082,
      "num_tokens": 14335385.0,
      "step": 7690
    },
    {
      "entropy": 5.9176054954528805,
      "epoch": 0.6795302013422819,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960614655623883,
      "loss": 5.6268,
      "mean_token_accuracy": 0.18449974656105042,
      "num_tokens": 14345005.0,
      "step": 7695
    },
    {
      "entropy": 5.917593383789063,
      "epoch": 0.6799717414341222,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496055596944233,
      "loss": 5.6335,
      "mean_token_accuracy": 0.1804521292448044,
      "num_tokens": 14354805.0,
      "step": 7700
    },
    {
      "entropy": 5.983846282958984,
      "epoch": 0.6804132815259626,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960497239957243,
      "loss": 5.6634,
      "mean_token_accuracy": 0.18017072528600692,
      "num_tokens": 14364798.0,
      "step": 7705
    },
    {
      "entropy": 5.964252614974976,
      "epoch": 0.6808548216178029,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960438467169772,
      "loss": 5.7475,
      "mean_token_accuracy": 0.1749880403280258,
      "num_tokens": 14374030.0,
      "step": 7710
    },
    {
      "entropy": 5.957871532440185,
      "epoch": 0.6812963617096433,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960379651081068,
      "loss": 5.6852,
      "mean_token_accuracy": 0.18233418762683867,
      "num_tokens": 14383859.0,
      "step": 7715
    },
    {
      "entropy": 5.986110353469849,
      "epoch": 0.6817379018014835,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960320791692283,
      "loss": 5.6304,
      "mean_token_accuracy": 0.17918013632297516,
      "num_tokens": 14392807.0,
      "step": 7720
    },
    {
      "entropy": 5.917441654205322,
      "epoch": 0.6821794418933239,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960261889004572,
      "loss": 5.6916,
      "mean_token_accuracy": 0.18358218669891357,
      "num_tokens": 14401135.0,
      "step": 7725
    },
    {
      "entropy": 5.998836660385132,
      "epoch": 0.6826209819851643,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004960202943019088,
      "loss": 5.6933,
      "mean_token_accuracy": 0.17721178084611894,
      "num_tokens": 14409811.0,
      "step": 7730
    },
    {
      "entropy": 5.987340831756592,
      "epoch": 0.6830625220770046,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960143953736984,
      "loss": 5.75,
      "mean_token_accuracy": 0.18547469824552537,
      "num_tokens": 14418639.0,
      "step": 7735
    },
    {
      "entropy": 5.957559490203858,
      "epoch": 0.683504062168845,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960084921159419,
      "loss": 5.6969,
      "mean_token_accuracy": 0.17965892553329468,
      "num_tokens": 14427980.0,
      "step": 7740
    },
    {
      "entropy": 5.899186849594116,
      "epoch": 0.6839456022606852,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960025845287546,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1840589702129364,
      "num_tokens": 14436926.0,
      "step": 7745
    },
    {
      "entropy": 5.966268253326416,
      "epoch": 0.6843871423525256,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959966726122523,
      "loss": 5.708,
      "mean_token_accuracy": 0.1810564011335373,
      "num_tokens": 14445945.0,
      "step": 7750
    },
    {
      "entropy": 5.899870681762695,
      "epoch": 0.6848286824443659,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000495990756366551,
      "loss": 5.6974,
      "mean_token_accuracy": 0.17831700593233107,
      "num_tokens": 14454727.0,
      "step": 7755
    },
    {
      "entropy": 5.928788805007935,
      "epoch": 0.6852702225362063,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959848357917664,
      "loss": 5.6774,
      "mean_token_accuracy": 0.18681171238422395,
      "num_tokens": 14463367.0,
      "step": 7760
    },
    {
      "entropy": 5.947967147827148,
      "epoch": 0.6857117626280467,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959789108880145,
      "loss": 5.5775,
      "mean_token_accuracy": 0.18846215903759003,
      "num_tokens": 14472144.0,
      "step": 7765
    },
    {
      "entropy": 6.044126749038696,
      "epoch": 0.686153302719887,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959729816554115,
      "loss": 5.774,
      "mean_token_accuracy": 0.17542928010225295,
      "num_tokens": 14481433.0,
      "step": 7770
    },
    {
      "entropy": 5.964986181259155,
      "epoch": 0.6865948428117273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004959670480940734,
      "loss": 5.6828,
      "mean_token_accuracy": 0.18336876332759858,
      "num_tokens": 14491081.0,
      "step": 7775
    },
    {
      "entropy": 5.906917095184326,
      "epoch": 0.6870363829035676,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959611102041165,
      "loss": 5.6828,
      "mean_token_accuracy": 0.18097760975360871,
      "num_tokens": 14500821.0,
      "step": 7780
    },
    {
      "entropy": 5.985409879684449,
      "epoch": 0.687477922995408,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004959551679856571,
      "loss": 5.6437,
      "mean_token_accuracy": 0.1880077987909317,
      "num_tokens": 14509591.0,
      "step": 7785
    },
    {
      "entropy": 5.971234226226807,
      "epoch": 0.6879194630872483,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959492214388117,
      "loss": 5.7331,
      "mean_token_accuracy": 0.1832096979022026,
      "num_tokens": 14519430.0,
      "step": 7790
    },
    {
      "entropy": 5.888854837417602,
      "epoch": 0.6883610031790887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004959432705636968,
      "loss": 5.6848,
      "mean_token_accuracy": 0.1821226879954338,
      "num_tokens": 14527321.0,
      "step": 7795
    },
    {
      "entropy": 5.923944997787475,
      "epoch": 0.688802543270929,
      "grad_norm": 1.0,
      "learning_rate": 0.0004959373153604287,
      "loss": 5.605,
      "mean_token_accuracy": 0.1885512426495552,
      "num_tokens": 14536579.0,
      "step": 7800
    },
    {
      "entropy": 5.943047094345093,
      "epoch": 0.6892440833627693,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959313558291243,
      "loss": 5.7668,
      "mean_token_accuracy": 0.18325393348932267,
      "num_tokens": 14547494.0,
      "step": 7805
    },
    {
      "entropy": 5.958702754974365,
      "epoch": 0.6896856234546097,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004959253919699004,
      "loss": 5.7584,
      "mean_token_accuracy": 0.17894975394010543,
      "num_tokens": 14557880.0,
      "step": 7810
    },
    {
      "entropy": 5.959938001632691,
      "epoch": 0.69012716354645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959194237828735,
      "loss": 5.6315,
      "mean_token_accuracy": 0.18977242410182954,
      "num_tokens": 14566838.0,
      "step": 7815
    },
    {
      "entropy": 5.918881273269653,
      "epoch": 0.6905687036382904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959134512681609,
      "loss": 5.5912,
      "mean_token_accuracy": 0.18764030635356904,
      "num_tokens": 14576039.0,
      "step": 7820
    },
    {
      "entropy": 5.872348022460938,
      "epoch": 0.6910102437301306,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959074744258794,
      "loss": 5.6604,
      "mean_token_accuracy": 0.18240880966186523,
      "num_tokens": 14585170.0,
      "step": 7825
    },
    {
      "entropy": 5.969253396987915,
      "epoch": 0.691451783821971,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959014932561461,
      "loss": 5.6427,
      "mean_token_accuracy": 0.18010437935590745,
      "num_tokens": 14594504.0,
      "step": 7830
    },
    {
      "entropy": 5.924509048461914,
      "epoch": 0.6918933239138114,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958955077590782,
      "loss": 5.7019,
      "mean_token_accuracy": 0.1792963683605194,
      "num_tokens": 14603571.0,
      "step": 7835
    },
    {
      "entropy": 5.912838506698608,
      "epoch": 0.6923348640056517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004958895179347929,
      "loss": 5.6356,
      "mean_token_accuracy": 0.18071549534797668,
      "num_tokens": 14612387.0,
      "step": 7840
    },
    {
      "entropy": 5.997235679626465,
      "epoch": 0.6927764040974921,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004958835237834075,
      "loss": 5.6689,
      "mean_token_accuracy": 0.18430794775485992,
      "num_tokens": 14620490.0,
      "step": 7845
    },
    {
      "entropy": 5.872093534469604,
      "epoch": 0.6932179441893324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004958775253050396,
      "loss": 5.6848,
      "mean_token_accuracy": 0.18527197688817978,
      "num_tokens": 14629592.0,
      "step": 7850
    },
    {
      "entropy": 5.988802051544189,
      "epoch": 0.6936594842811727,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958715224998066,
      "loss": 5.7043,
      "mean_token_accuracy": 0.17896327674388884,
      "num_tokens": 14639358.0,
      "step": 7855
    },
    {
      "entropy": 6.046493053436279,
      "epoch": 0.694101024373013,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004958655153678261,
      "loss": 5.7201,
      "mean_token_accuracy": 0.17660781145095825,
      "num_tokens": 14649046.0,
      "step": 7860
    },
    {
      "entropy": 5.919287919998169,
      "epoch": 0.6945425644648534,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004958595039092156,
      "loss": 5.7001,
      "mean_token_accuracy": 0.18351273387670516,
      "num_tokens": 14658395.0,
      "step": 7865
    },
    {
      "entropy": 5.939451122283936,
      "epoch": 0.6949841045566938,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958534881240932,
      "loss": 5.7359,
      "mean_token_accuracy": 0.17813275754451752,
      "num_tokens": 14668886.0,
      "step": 7870
    },
    {
      "entropy": 5.980077886581421,
      "epoch": 0.6954256446485341,
      "grad_norm": 1.0,
      "learning_rate": 0.0004958474680125766,
      "loss": 5.5826,
      "mean_token_accuracy": 0.18360820710659026,
      "num_tokens": 14676683.0,
      "step": 7875
    },
    {
      "entropy": 5.896993732452392,
      "epoch": 0.6958671847403745,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004958414435747837,
      "loss": 5.6899,
      "mean_token_accuracy": 0.18220269829034805,
      "num_tokens": 14685902.0,
      "step": 7880
    },
    {
      "entropy": 5.858180570602417,
      "epoch": 0.6963087248322147,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004958354148108324,
      "loss": 5.5683,
      "mean_token_accuracy": 0.1955101802945137,
      "num_tokens": 14695215.0,
      "step": 7885
    },
    {
      "entropy": 5.9535950183868405,
      "epoch": 0.6967502649240551,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495829381720841,
      "loss": 5.7295,
      "mean_token_accuracy": 0.1743612363934517,
      "num_tokens": 14703458.0,
      "step": 7890
    },
    {
      "entropy": 5.9401897430419925,
      "epoch": 0.6971918050158954,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004958233443049275,
      "loss": 5.6325,
      "mean_token_accuracy": 0.18294671326875686,
      "num_tokens": 14712298.0,
      "step": 7895
    },
    {
      "entropy": 5.9343496799469,
      "epoch": 0.6976333451077358,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958173025632103,
      "loss": 5.6762,
      "mean_token_accuracy": 0.17693610191345216,
      "num_tokens": 14721405.0,
      "step": 7900
    },
    {
      "entropy": 5.94487681388855,
      "epoch": 0.6980748851995762,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004958112564958079,
      "loss": 5.703,
      "mean_token_accuracy": 0.18319268971681596,
      "num_tokens": 14731571.0,
      "step": 7905
    },
    {
      "entropy": 5.916205501556396,
      "epoch": 0.6985164252914164,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958052061028385,
      "loss": 5.6826,
      "mean_token_accuracy": 0.1873116374015808,
      "num_tokens": 14739964.0,
      "step": 7910
    },
    {
      "entropy": 5.965706443786621,
      "epoch": 0.6989579653832568,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957991513844205,
      "loss": 5.7819,
      "mean_token_accuracy": 0.173875692486763,
      "num_tokens": 14749283.0,
      "step": 7915
    },
    {
      "entropy": 6.055011129379272,
      "epoch": 0.6993995054750971,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004957930923406729,
      "loss": 5.9137,
      "mean_token_accuracy": 0.1639314651489258,
      "num_tokens": 14759973.0,
      "step": 7920
    },
    {
      "entropy": 6.058173799514771,
      "epoch": 0.6998410455669375,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004957870289717142,
      "loss": 5.7777,
      "mean_token_accuracy": 0.1704869329929352,
      "num_tokens": 14768648.0,
      "step": 7925
    },
    {
      "entropy": 6.0284524917602536,
      "epoch": 0.7002825856587778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004957809612776631,
      "loss": 5.8041,
      "mean_token_accuracy": 0.1774164095520973,
      "num_tokens": 14778022.0,
      "step": 7930
    },
    {
      "entropy": 6.028066253662109,
      "epoch": 0.7007241257506182,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957748892586384,
      "loss": 5.7865,
      "mean_token_accuracy": 0.17293851673603058,
      "num_tokens": 14787949.0,
      "step": 7935
    },
    {
      "entropy": 5.8997344970703125,
      "epoch": 0.7011656658424585,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957688129147594,
      "loss": 5.6904,
      "mean_token_accuracy": 0.1884993314743042,
      "num_tokens": 14796862.0,
      "step": 7940
    },
    {
      "entropy": 5.967758226394653,
      "epoch": 0.7016072059342988,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957627322461448,
      "loss": 5.7214,
      "mean_token_accuracy": 0.17945120632648467,
      "num_tokens": 14805824.0,
      "step": 7945
    },
    {
      "entropy": 5.9464047908782955,
      "epoch": 0.7020487460261392,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957566472529139,
      "loss": 5.697,
      "mean_token_accuracy": 0.18021153509616852,
      "num_tokens": 14815322.0,
      "step": 7950
    },
    {
      "entropy": 5.985154867172241,
      "epoch": 0.7024902861179795,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957505579351858,
      "loss": 5.7429,
      "mean_token_accuracy": 0.16869543343782425,
      "num_tokens": 14825116.0,
      "step": 7955
    },
    {
      "entropy": 5.923246288299561,
      "epoch": 0.7029318262098199,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004957444642930798,
      "loss": 5.6286,
      "mean_token_accuracy": 0.1803962543606758,
      "num_tokens": 14834011.0,
      "step": 7960
    },
    {
      "entropy": 5.804202318191528,
      "epoch": 0.7033733663016601,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957383663267152,
      "loss": 5.5152,
      "mean_token_accuracy": 0.19704827964305877,
      "num_tokens": 14842448.0,
      "step": 7965
    },
    {
      "entropy": 5.941469764709472,
      "epoch": 0.7038149063935005,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957322640362118,
      "loss": 5.7336,
      "mean_token_accuracy": 0.17948463410139084,
      "num_tokens": 14851050.0,
      "step": 7970
    },
    {
      "entropy": 6.022776079177857,
      "epoch": 0.7042564464853409,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957261574216887,
      "loss": 5.7191,
      "mean_token_accuracy": 0.1809219777584076,
      "num_tokens": 14860240.0,
      "step": 7975
    },
    {
      "entropy": 5.967269515991211,
      "epoch": 0.7046979865771812,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957200464832656,
      "loss": 5.6233,
      "mean_token_accuracy": 0.19388763904571532,
      "num_tokens": 14869835.0,
      "step": 7980
    },
    {
      "entropy": 5.966859483718872,
      "epoch": 0.7051395266690216,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957139312210626,
      "loss": 5.6648,
      "mean_token_accuracy": 0.18310383558273316,
      "num_tokens": 14878907.0,
      "step": 7985
    },
    {
      "entropy": 5.830237197875976,
      "epoch": 0.7055810667608619,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495707811635199,
      "loss": 5.5936,
      "mean_token_accuracy": 0.1895356595516205,
      "num_tokens": 14887368.0,
      "step": 7990
    },
    {
      "entropy": 5.915153312683105,
      "epoch": 0.7060226068527022,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004957016877257949,
      "loss": 5.6654,
      "mean_token_accuracy": 0.1790101557970047,
      "num_tokens": 14896095.0,
      "step": 7995
    },
    {
      "entropy": 6.018905162811279,
      "epoch": 0.7064641469445425,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956955594929704,
      "loss": 5.6396,
      "mean_token_accuracy": 0.18453716039657592,
      "num_tokens": 14904486.0,
      "step": 8000
    },
    {
      "entropy": 5.943782949447632,
      "epoch": 0.7069056870363829,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956894269368454,
      "loss": 5.7492,
      "mean_token_accuracy": 0.17737880498170852,
      "num_tokens": 14914987.0,
      "step": 8005
    },
    {
      "entropy": 6.0336995124816895,
      "epoch": 0.7073472271282233,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956832900575399,
      "loss": 5.786,
      "mean_token_accuracy": 0.16838618367910385,
      "num_tokens": 14924283.0,
      "step": 8010
    },
    {
      "entropy": 5.99911322593689,
      "epoch": 0.7077887672200636,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004956771488551742,
      "loss": 5.689,
      "mean_token_accuracy": 0.18882422745227814,
      "num_tokens": 14933116.0,
      "step": 8015
    },
    {
      "entropy": 5.953978872299194,
      "epoch": 0.708230307311904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956710033298689,
      "loss": 5.6579,
      "mean_token_accuracy": 0.18190203756093978,
      "num_tokens": 14941684.0,
      "step": 8020
    },
    {
      "entropy": 5.848041152954101,
      "epoch": 0.7086718474037442,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495664853481744,
      "loss": 5.6435,
      "mean_token_accuracy": 0.1871766746044159,
      "num_tokens": 14951005.0,
      "step": 8025
    },
    {
      "entropy": 5.975212526321411,
      "epoch": 0.7091133874955846,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956586993109201,
      "loss": 5.7044,
      "mean_token_accuracy": 0.18854624629020691,
      "num_tokens": 14960226.0,
      "step": 8030
    },
    {
      "entropy": 5.983984899520874,
      "epoch": 0.7095549275874249,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956525408175179,
      "loss": 5.806,
      "mean_token_accuracy": 0.171312353014946,
      "num_tokens": 14970022.0,
      "step": 8035
    },
    {
      "entropy": 6.001506519317627,
      "epoch": 0.7099964676792653,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956463780016578,
      "loss": 5.7418,
      "mean_token_accuracy": 0.1778755769133568,
      "num_tokens": 14978855.0,
      "step": 8040
    },
    {
      "entropy": 5.980483055114746,
      "epoch": 0.7104380077711057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004956402108634607,
      "loss": 5.6487,
      "mean_token_accuracy": 0.18839399814605712,
      "num_tokens": 14987299.0,
      "step": 8045
    },
    {
      "entropy": 5.931147336959839,
      "epoch": 0.7108795478629459,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004956340394030473,
      "loss": 5.7713,
      "mean_token_accuracy": 0.17761172652244567,
      "num_tokens": 14997330.0,
      "step": 8050
    },
    {
      "entropy": 6.0416614532470705,
      "epoch": 0.7113210879547863,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004956278636205386,
      "loss": 5.7518,
      "mean_token_accuracy": 0.17855996638536453,
      "num_tokens": 15007194.0,
      "step": 8055
    },
    {
      "entropy": 5.965719079971313,
      "epoch": 0.7117626280466266,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956216835160556,
      "loss": 5.76,
      "mean_token_accuracy": 0.17949917912483215,
      "num_tokens": 15017349.0,
      "step": 8060
    },
    {
      "entropy": 5.976887893676758,
      "epoch": 0.712204168138467,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004956154990897191,
      "loss": 5.7607,
      "mean_token_accuracy": 0.1701971098780632,
      "num_tokens": 15026788.0,
      "step": 8065
    },
    {
      "entropy": 6.012468242645264,
      "epoch": 0.7126457082303073,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004956093103416505,
      "loss": 5.7185,
      "mean_token_accuracy": 0.1757481426000595,
      "num_tokens": 15034656.0,
      "step": 8070
    },
    {
      "entropy": 5.906048059463501,
      "epoch": 0.7130872483221476,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956031172719711,
      "loss": 5.641,
      "mean_token_accuracy": 0.1795785665512085,
      "num_tokens": 15043849.0,
      "step": 8075
    },
    {
      "entropy": 5.984164667129517,
      "epoch": 0.713528788413988,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955969198808021,
      "loss": 5.6707,
      "mean_token_accuracy": 0.183877694606781,
      "num_tokens": 15052834.0,
      "step": 8080
    },
    {
      "entropy": 5.97470235824585,
      "epoch": 0.7139703285058283,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955907181682648,
      "loss": 5.7793,
      "mean_token_accuracy": 0.1751793786883354,
      "num_tokens": 15062782.0,
      "step": 8085
    },
    {
      "entropy": 5.933545017242432,
      "epoch": 0.7144118685976687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955845121344809,
      "loss": 5.6041,
      "mean_token_accuracy": 0.19034629315137863,
      "num_tokens": 15071456.0,
      "step": 8090
    },
    {
      "entropy": 5.976210880279541,
      "epoch": 0.714853408689509,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955783017795717,
      "loss": 5.6517,
      "mean_token_accuracy": 0.1833293244242668,
      "num_tokens": 15080256.0,
      "step": 8095
    },
    {
      "entropy": 5.963651895523071,
      "epoch": 0.7152949487813494,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000495572087103659,
      "loss": 5.7492,
      "mean_token_accuracy": 0.17811937183141707,
      "num_tokens": 15089541.0,
      "step": 8100
    },
    {
      "entropy": 5.921456384658813,
      "epoch": 0.7157364888731896,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955658681068647,
      "loss": 5.6077,
      "mean_token_accuracy": 0.1819453254342079,
      "num_tokens": 15097853.0,
      "step": 8105
    },
    {
      "entropy": 5.943249607086182,
      "epoch": 0.71617802896503,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004955596447893105,
      "loss": 5.6641,
      "mean_token_accuracy": 0.1846989244222641,
      "num_tokens": 15107816.0,
      "step": 8110
    },
    {
      "entropy": 6.019874334335327,
      "epoch": 0.7166195690568704,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955534171511181,
      "loss": 5.728,
      "mean_token_accuracy": 0.17390682250261308,
      "num_tokens": 15116823.0,
      "step": 8115
    },
    {
      "entropy": 5.9666664600372314,
      "epoch": 0.7170611091487107,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00049554718519241,
      "loss": 5.727,
      "mean_token_accuracy": 0.18093488961458207,
      "num_tokens": 15126479.0,
      "step": 8120
    },
    {
      "entropy": 5.924319410324097,
      "epoch": 0.7175026492405511,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004955409489133078,
      "loss": 5.5786,
      "mean_token_accuracy": 0.18755926489830016,
      "num_tokens": 15134830.0,
      "step": 8125
    },
    {
      "entropy": 5.939011430740356,
      "epoch": 0.7179441893323913,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004955347083139338,
      "loss": 5.6547,
      "mean_token_accuracy": 0.19012102633714675,
      "num_tokens": 15143871.0,
      "step": 8130
    },
    {
      "entropy": 5.964648485183716,
      "epoch": 0.7183857294242317,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955284633944104,
      "loss": 5.7296,
      "mean_token_accuracy": 0.17857578843832017,
      "num_tokens": 15153378.0,
      "step": 8135
    },
    {
      "entropy": 5.879282379150391,
      "epoch": 0.718827269516072,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955222141548597,
      "loss": 5.614,
      "mean_token_accuracy": 0.19564612656831742,
      "num_tokens": 15163238.0,
      "step": 8140
    },
    {
      "entropy": 6.047828197479248,
      "epoch": 0.7192688096079124,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955159605954043,
      "loss": 5.7633,
      "mean_token_accuracy": 0.1803179532289505,
      "num_tokens": 15172280.0,
      "step": 8145
    },
    {
      "entropy": 5.900742816925049,
      "epoch": 0.7197103496997528,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004955097027161666,
      "loss": 5.6312,
      "mean_token_accuracy": 0.176329106092453,
      "num_tokens": 15181687.0,
      "step": 8150
    },
    {
      "entropy": 5.887002229690552,
      "epoch": 0.7201518897915931,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004955034405172693,
      "loss": 5.6454,
      "mean_token_accuracy": 0.18313996940851213,
      "num_tokens": 15190644.0,
      "step": 8155
    },
    {
      "entropy": 5.977613115310669,
      "epoch": 0.7205934298834334,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495497173998835,
      "loss": 5.7866,
      "mean_token_accuracy": 0.17292434126138687,
      "num_tokens": 15201159.0,
      "step": 8160
    },
    {
      "entropy": 5.97907509803772,
      "epoch": 0.7210349699752737,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004954909031609864,
      "loss": 5.6803,
      "mean_token_accuracy": 0.17891637980937958,
      "num_tokens": 15210351.0,
      "step": 8165
    },
    {
      "entropy": 5.943047666549683,
      "epoch": 0.7214765100671141,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004954846280038465,
      "loss": 5.7457,
      "mean_token_accuracy": 0.18264581114053727,
      "num_tokens": 15219022.0,
      "step": 8170
    },
    {
      "entropy": 5.970562744140625,
      "epoch": 0.7219180501589544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004954783485275381,
      "loss": 5.7098,
      "mean_token_accuracy": 0.17534444332122803,
      "num_tokens": 15228274.0,
      "step": 8175
    },
    {
      "entropy": 6.010308694839478,
      "epoch": 0.7223595902507948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954720647321843,
      "loss": 5.7362,
      "mean_token_accuracy": 0.17584520429372788,
      "num_tokens": 15238079.0,
      "step": 8180
    },
    {
      "entropy": 5.946929836273194,
      "epoch": 0.7228011303426352,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004954657766179081,
      "loss": 5.5626,
      "mean_token_accuracy": 0.1946161523461342,
      "num_tokens": 15247313.0,
      "step": 8185
    },
    {
      "entropy": 5.926880121231079,
      "epoch": 0.7232426704344754,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954594841848327,
      "loss": 5.6389,
      "mean_token_accuracy": 0.19094515591859818,
      "num_tokens": 15255964.0,
      "step": 8190
    },
    {
      "entropy": 5.963173580169678,
      "epoch": 0.7236842105263158,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004954531874330814,
      "loss": 5.6726,
      "mean_token_accuracy": 0.1774781733751297,
      "num_tokens": 15265113.0,
      "step": 8195
    },
    {
      "entropy": 5.931451606750488,
      "epoch": 0.7241257506181561,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004954468863627774,
      "loss": 5.7006,
      "mean_token_accuracy": 0.1804114207625389,
      "num_tokens": 15275324.0,
      "step": 8200
    },
    {
      "entropy": 5.902015733718872,
      "epoch": 0.7245672907099965,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954405809740445,
      "loss": 5.599,
      "mean_token_accuracy": 0.18788865059614182,
      "num_tokens": 15284309.0,
      "step": 8205
    },
    {
      "entropy": 5.915576457977295,
      "epoch": 0.7250088308018368,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004954342712670058,
      "loss": 5.6467,
      "mean_token_accuracy": 0.18402685821056367,
      "num_tokens": 15293090.0,
      "step": 8210
    },
    {
      "entropy": 5.888005495071411,
      "epoch": 0.7254503708936771,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954279572417852,
      "loss": 5.5629,
      "mean_token_accuracy": 0.20007124841213225,
      "num_tokens": 15302351.0,
      "step": 8215
    },
    {
      "entropy": 5.974038028717041,
      "epoch": 0.7258919109855175,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004954216388985063,
      "loss": 5.7139,
      "mean_token_accuracy": 0.1835293173789978,
      "num_tokens": 15311232.0,
      "step": 8220
    },
    {
      "entropy": 5.891401863098144,
      "epoch": 0.7263334510773578,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954153162372928,
      "loss": 5.6977,
      "mean_token_accuracy": 0.17834204584360122,
      "num_tokens": 15320096.0,
      "step": 8225
    },
    {
      "entropy": 5.929154014587402,
      "epoch": 0.7267749911691982,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954089892582686,
      "loss": 5.6235,
      "mean_token_accuracy": 0.1872919261455536,
      "num_tokens": 15328754.0,
      "step": 8230
    },
    {
      "entropy": 5.990077257156372,
      "epoch": 0.7272165312610385,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004954026579615577,
      "loss": 5.7356,
      "mean_token_accuracy": 0.17519152611494065,
      "num_tokens": 15338315.0,
      "step": 8235
    },
    {
      "entropy": 5.959997034072876,
      "epoch": 0.7276580713528789,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953963223472841,
      "loss": 5.7156,
      "mean_token_accuracy": 0.1840008869767189,
      "num_tokens": 15347661.0,
      "step": 8240
    },
    {
      "entropy": 5.949524736404419,
      "epoch": 0.7280996114447191,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004953899824155718,
      "loss": 5.7363,
      "mean_token_accuracy": 0.17726923674345016,
      "num_tokens": 15357331.0,
      "step": 8245
    },
    {
      "entropy": 5.958864736557007,
      "epoch": 0.7285411515365595,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953836381665452,
      "loss": 5.6491,
      "mean_token_accuracy": 0.1837444558739662,
      "num_tokens": 15365904.0,
      "step": 8250
    },
    {
      "entropy": 5.858493328094482,
      "epoch": 0.7289826916283999,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953772896003284,
      "loss": 5.5945,
      "mean_token_accuracy": 0.18516038060188295,
      "num_tokens": 15375745.0,
      "step": 8255
    },
    {
      "entropy": 5.951942491531372,
      "epoch": 0.7294242317202402,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953709367170458,
      "loss": 5.6526,
      "mean_token_accuracy": 0.1849004715681076,
      "num_tokens": 15384745.0,
      "step": 8260
    },
    {
      "entropy": 5.970254945755005,
      "epoch": 0.7298657718120806,
      "grad_norm": 0.96875,
      "learning_rate": 0.000495364579516822,
      "loss": 5.7297,
      "mean_token_accuracy": 0.17167257219552995,
      "num_tokens": 15393300.0,
      "step": 8265
    },
    {
      "entropy": 5.904949760437011,
      "epoch": 0.7303073119039208,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004953582179997813,
      "loss": 5.7199,
      "mean_token_accuracy": 0.17883816212415696,
      "num_tokens": 15402689.0,
      "step": 8270
    },
    {
      "entropy": 6.024151563644409,
      "epoch": 0.7307488519957612,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004953518521660484,
      "loss": 5.7722,
      "mean_token_accuracy": 0.17414382100105286,
      "num_tokens": 15412513.0,
      "step": 8275
    },
    {
      "entropy": 6.0064616203308105,
      "epoch": 0.7311903920876015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004953454820157481,
      "loss": 5.6428,
      "mean_token_accuracy": 0.18268238008022308,
      "num_tokens": 15421639.0,
      "step": 8280
    },
    {
      "entropy": 5.937090444564819,
      "epoch": 0.7316319321794419,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953391075490051,
      "loss": 5.712,
      "mean_token_accuracy": 0.18467180728912352,
      "num_tokens": 15430361.0,
      "step": 8285
    },
    {
      "entropy": 5.945833015441894,
      "epoch": 0.7320734722712823,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953327287659444,
      "loss": 5.7323,
      "mean_token_accuracy": 0.18001717925071717,
      "num_tokens": 15440262.0,
      "step": 8290
    },
    {
      "entropy": 5.9669712543487545,
      "epoch": 0.7325150123631226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004953263456666907,
      "loss": 5.6507,
      "mean_token_accuracy": 0.1816618725657463,
      "num_tokens": 15449355.0,
      "step": 8295
    },
    {
      "entropy": 5.9731669425964355,
      "epoch": 0.7329565524549629,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004953199582513693,
      "loss": 5.6645,
      "mean_token_accuracy": 0.1790527105331421,
      "num_tokens": 15458482.0,
      "step": 8300
    },
    {
      "entropy": 5.97604718208313,
      "epoch": 0.7333980925468032,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004953135665201052,
      "loss": 5.7017,
      "mean_token_accuracy": 0.17600667476654053,
      "num_tokens": 15468243.0,
      "step": 8305
    },
    {
      "entropy": 5.94189338684082,
      "epoch": 0.7338396326386436,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004953071704730237,
      "loss": 5.663,
      "mean_token_accuracy": 0.18367537707090378,
      "num_tokens": 15477674.0,
      "step": 8310
    },
    {
      "entropy": 5.946559715270996,
      "epoch": 0.7342811727304839,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004953007701102499,
      "loss": 5.6983,
      "mean_token_accuracy": 0.18252862691879274,
      "num_tokens": 15487456.0,
      "step": 8315
    },
    {
      "entropy": 5.9820044994354244,
      "epoch": 0.7347227128223243,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004952943654319094,
      "loss": 5.7952,
      "mean_token_accuracy": 0.174931663274765,
      "num_tokens": 15497541.0,
      "step": 8320
    },
    {
      "entropy": 6.0301002025604244,
      "epoch": 0.7351642529141647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952879564381276,
      "loss": 5.6879,
      "mean_token_accuracy": 0.18314930200576782,
      "num_tokens": 15507142.0,
      "step": 8325
    },
    {
      "entropy": 5.905955648422241,
      "epoch": 0.7356057930060049,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952815431290298,
      "loss": 5.6591,
      "mean_token_accuracy": 0.18085090517997743,
      "num_tokens": 15516826.0,
      "step": 8330
    },
    {
      "entropy": 5.975840473175049,
      "epoch": 0.7360473330978453,
      "grad_norm": 1.0,
      "learning_rate": 0.000495275125504742,
      "loss": 5.8632,
      "mean_token_accuracy": 0.17261759787797928,
      "num_tokens": 15527123.0,
      "step": 8335
    },
    {
      "entropy": 6.067738342285156,
      "epoch": 0.7364888731896856,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004952687035653899,
      "loss": 5.7584,
      "mean_token_accuracy": 0.171932914853096,
      "num_tokens": 15536874.0,
      "step": 8340
    },
    {
      "entropy": 5.926822471618652,
      "epoch": 0.736930413281526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004952622773110989,
      "loss": 5.646,
      "mean_token_accuracy": 0.18517315983772278,
      "num_tokens": 15545585.0,
      "step": 8345
    },
    {
      "entropy": 6.0062745094299315,
      "epoch": 0.7373719533733663,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952558467419953,
      "loss": 5.7549,
      "mean_token_accuracy": 0.17019776701927186,
      "num_tokens": 15555089.0,
      "step": 8350
    },
    {
      "entropy": 5.992174816131592,
      "epoch": 0.7378134934652066,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000495249411858205,
      "loss": 5.7894,
      "mean_token_accuracy": 0.1737505778670311,
      "num_tokens": 15565630.0,
      "step": 8355
    },
    {
      "entropy": 5.934722661972046,
      "epoch": 0.738255033557047,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952429726598538,
      "loss": 5.6589,
      "mean_token_accuracy": 0.1861184924840927,
      "num_tokens": 15574898.0,
      "step": 8360
    },
    {
      "entropy": 5.990200424194336,
      "epoch": 0.7386965736488873,
      "grad_norm": 1.0,
      "learning_rate": 0.0004952365291470682,
      "loss": 5.7446,
      "mean_token_accuracy": 0.18089302778244018,
      "num_tokens": 15585633.0,
      "step": 8365
    },
    {
      "entropy": 5.978502225875855,
      "epoch": 0.7391381137407277,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952300813199742,
      "loss": 5.6587,
      "mean_token_accuracy": 0.18533686101436614,
      "num_tokens": 15594143.0,
      "step": 8370
    },
    {
      "entropy": 5.837214565277099,
      "epoch": 0.739579653832568,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952236291786981,
      "loss": 5.622,
      "mean_token_accuracy": 0.1932321935892105,
      "num_tokens": 15602812.0,
      "step": 8375
    },
    {
      "entropy": 5.905216026306152,
      "epoch": 0.7400211939244083,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952171727233665,
      "loss": 5.6268,
      "mean_token_accuracy": 0.1841863915324211,
      "num_tokens": 15612346.0,
      "step": 8380
    },
    {
      "entropy": 6.008640956878662,
      "epoch": 0.7404627340162486,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952107119541055,
      "loss": 5.7837,
      "mean_token_accuracy": 0.17194932848215103,
      "num_tokens": 15623064.0,
      "step": 8385
    },
    {
      "entropy": 5.967648029327393,
      "epoch": 0.740904274108089,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952042468710421,
      "loss": 5.69,
      "mean_token_accuracy": 0.1835721418261528,
      "num_tokens": 15631118.0,
      "step": 8390
    },
    {
      "entropy": 5.965255403518677,
      "epoch": 0.7413458141999294,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004951977774743027,
      "loss": 5.7025,
      "mean_token_accuracy": 0.17624314576387407,
      "num_tokens": 15641304.0,
      "step": 8395
    },
    {
      "entropy": 5.905878782272339,
      "epoch": 0.7417873542917697,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004951913037640139,
      "loss": 5.6168,
      "mean_token_accuracy": 0.18580237925052642,
      "num_tokens": 15649429.0,
      "step": 8400
    },
    {
      "entropy": 5.873190832138062,
      "epoch": 0.7422288943836101,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951848257403029,
      "loss": 5.6654,
      "mean_token_accuracy": 0.1869506448507309,
      "num_tokens": 15658577.0,
      "step": 8405
    },
    {
      "entropy": 5.956317949295044,
      "epoch": 0.7426704344754503,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951783434032963,
      "loss": 5.6022,
      "mean_token_accuracy": 0.19225453734397888,
      "num_tokens": 15667497.0,
      "step": 8410
    },
    {
      "entropy": 5.955981016159058,
      "epoch": 0.7431119745672907,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004951718567531211,
      "loss": 5.7012,
      "mean_token_accuracy": 0.16861613243818283,
      "num_tokens": 15677271.0,
      "step": 8415
    },
    {
      "entropy": 5.91002082824707,
      "epoch": 0.743553514659131,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951653657899045,
      "loss": 5.6527,
      "mean_token_accuracy": 0.1830442726612091,
      "num_tokens": 15686729.0,
      "step": 8420
    },
    {
      "entropy": 5.932142114639282,
      "epoch": 0.7439950547509714,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951588705137737,
      "loss": 5.6871,
      "mean_token_accuracy": 0.18392641842365265,
      "num_tokens": 15696229.0,
      "step": 8425
    },
    {
      "entropy": 5.922904968261719,
      "epoch": 0.7444365948428118,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004951523709248557,
      "loss": 5.6554,
      "mean_token_accuracy": 0.18072777539491652,
      "num_tokens": 15705631.0,
      "step": 8430
    },
    {
      "entropy": 5.914482450485229,
      "epoch": 0.744878134934652,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495145867023278,
      "loss": 5.5739,
      "mean_token_accuracy": 0.19226114600896835,
      "num_tokens": 15713953.0,
      "step": 8435
    },
    {
      "entropy": 6.018058633804321,
      "epoch": 0.7453196750264924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495139358809168,
      "loss": 5.8041,
      "mean_token_accuracy": 0.1720973879098892,
      "num_tokens": 15723885.0,
      "step": 8440
    },
    {
      "entropy": 5.969194078445435,
      "epoch": 0.7457612151183327,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004951328462826532,
      "loss": 5.7248,
      "mean_token_accuracy": 0.17739588171243667,
      "num_tokens": 15733879.0,
      "step": 8445
    },
    {
      "entropy": 5.935660409927368,
      "epoch": 0.7462027552101731,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951263294438611,
      "loss": 5.678,
      "mean_token_accuracy": 0.18263277411460876,
      "num_tokens": 15743177.0,
      "step": 8450
    },
    {
      "entropy": 5.964620923995971,
      "epoch": 0.7466442953020134,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951198082929194,
      "loss": 5.6946,
      "mean_token_accuracy": 0.1766124799847603,
      "num_tokens": 15751928.0,
      "step": 8455
    },
    {
      "entropy": 5.948552989959717,
      "epoch": 0.7470858353938538,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951132828299558,
      "loss": 5.7273,
      "mean_token_accuracy": 0.17715592980384826,
      "num_tokens": 15760950.0,
      "step": 8460
    },
    {
      "entropy": 5.973338794708252,
      "epoch": 0.7475273754856941,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004951067530550982,
      "loss": 5.7282,
      "mean_token_accuracy": 0.17724989503622054,
      "num_tokens": 15771099.0,
      "step": 8465
    },
    {
      "entropy": 5.9191803455352785,
      "epoch": 0.7479689155775344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951002189684745,
      "loss": 5.6525,
      "mean_token_accuracy": 0.17876300364732742,
      "num_tokens": 15781146.0,
      "step": 8470
    },
    {
      "entropy": 6.018130588531494,
      "epoch": 0.7484104556693748,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950936805702127,
      "loss": 5.7788,
      "mean_token_accuracy": 0.1711433693766594,
      "num_tokens": 15790817.0,
      "step": 8475
    },
    {
      "entropy": 6.004199457168579,
      "epoch": 0.7488519957612151,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004950871378604409,
      "loss": 5.7424,
      "mean_token_accuracy": 0.1754971355199814,
      "num_tokens": 15800160.0,
      "step": 8480
    },
    {
      "entropy": 5.957149887084961,
      "epoch": 0.7492935358530555,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950805908392872,
      "loss": 5.6911,
      "mean_token_accuracy": 0.1819565400481224,
      "num_tokens": 15809299.0,
      "step": 8485
    },
    {
      "entropy": 6.003140544891357,
      "epoch": 0.7497350759448957,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950740395068799,
      "loss": 5.7117,
      "mean_token_accuracy": 0.18326081782579423,
      "num_tokens": 15818874.0,
      "step": 8490
    },
    {
      "entropy": 5.9831794738769535,
      "epoch": 0.7501766160367361,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004950674838633473,
      "loss": 5.7774,
      "mean_token_accuracy": 0.1703500360250473,
      "num_tokens": 15828442.0,
      "step": 8495
    },
    {
      "entropy": 5.971370983123779,
      "epoch": 0.7506181561285765,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004950609239088178,
      "loss": 5.6698,
      "mean_token_accuracy": 0.1847639948129654,
      "num_tokens": 15838011.0,
      "step": 8500
    },
    {
      "entropy": 5.958171701431274,
      "epoch": 0.7510596962204168,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004950543596434201,
      "loss": 5.6556,
      "mean_token_accuracy": 0.1874200478196144,
      "num_tokens": 15846217.0,
      "step": 8505
    },
    {
      "entropy": 5.930761623382568,
      "epoch": 0.7515012363122572,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004950477910672825,
      "loss": 5.6556,
      "mean_token_accuracy": 0.1844528615474701,
      "num_tokens": 15855315.0,
      "step": 8510
    },
    {
      "entropy": 5.901986837387085,
      "epoch": 0.7519427764040975,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495041218180534,
      "loss": 5.6048,
      "mean_token_accuracy": 0.18934593349695206,
      "num_tokens": 15863150.0,
      "step": 8515
    },
    {
      "entropy": 5.879916429519653,
      "epoch": 0.7523843164959378,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495034640983303,
      "loss": 5.5811,
      "mean_token_accuracy": 0.18448757976293564,
      "num_tokens": 15871898.0,
      "step": 8520
    },
    {
      "entropy": 5.997114610671997,
      "epoch": 0.7528258565877781,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950280594757186,
      "loss": 5.698,
      "mean_token_accuracy": 0.18438488394021987,
      "num_tokens": 15880630.0,
      "step": 8525
    },
    {
      "entropy": 5.900802278518677,
      "epoch": 0.7532673966796185,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950214736579097,
      "loss": 5.6454,
      "mean_token_accuracy": 0.187907612323761,
      "num_tokens": 15889798.0,
      "step": 8530
    },
    {
      "entropy": 5.933571434020996,
      "epoch": 0.7537089367714589,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950148835300051,
      "loss": 5.676,
      "mean_token_accuracy": 0.187706558406353,
      "num_tokens": 15899123.0,
      "step": 8535
    },
    {
      "entropy": 6.00587739944458,
      "epoch": 0.7541504768632992,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950082890921341,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1778817743062973,
      "num_tokens": 15908856.0,
      "step": 8540
    },
    {
      "entropy": 5.876418590545654,
      "epoch": 0.7545920169551396,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950016903444261,
      "loss": 5.6074,
      "mean_token_accuracy": 0.19075658768415452,
      "num_tokens": 15918983.0,
      "step": 8545
    },
    {
      "entropy": 5.85601224899292,
      "epoch": 0.7550335570469798,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949950872870099,
      "loss": 5.6264,
      "mean_token_accuracy": 0.18571536839008332,
      "num_tokens": 15927764.0,
      "step": 8550
    },
    {
      "entropy": 5.949892044067383,
      "epoch": 0.7554750971388202,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494988479920015,
      "loss": 5.6595,
      "mean_token_accuracy": 0.1836940437555313,
      "num_tokens": 15936497.0,
      "step": 8555
    },
    {
      "entropy": 5.9826719760894775,
      "epoch": 0.7559166372306605,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494981868243571,
      "loss": 5.6129,
      "mean_token_accuracy": 0.19229195415973663,
      "num_tokens": 15946151.0,
      "step": 8560
    },
    {
      "entropy": 5.95530219078064,
      "epoch": 0.7563581773225009,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949752522578073,
      "loss": 5.6482,
      "mean_token_accuracy": 0.19404937326908112,
      "num_tokens": 15955278.0,
      "step": 8565
    },
    {
      "entropy": 5.948098278045654,
      "epoch": 0.7567997174143413,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949686319628535,
      "loss": 5.6458,
      "mean_token_accuracy": 0.1796131893992424,
      "num_tokens": 15964260.0,
      "step": 8570
    },
    {
      "entropy": 5.9679864883422855,
      "epoch": 0.7572412575061815,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004949620073588394,
      "loss": 5.7201,
      "mean_token_accuracy": 0.1845954105257988,
      "num_tokens": 15973531.0,
      "step": 8575
    },
    {
      "entropy": 5.876770925521851,
      "epoch": 0.7576827975980219,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004949553784458945,
      "loss": 5.5916,
      "mean_token_accuracy": 0.1945582002401352,
      "num_tokens": 15981602.0,
      "step": 8580
    },
    {
      "entropy": 5.9718015670776365,
      "epoch": 0.7581243376898622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004949487452241489,
      "loss": 5.6924,
      "mean_token_accuracy": 0.1735656216740608,
      "num_tokens": 15991114.0,
      "step": 8585
    },
    {
      "entropy": 5.92060546875,
      "epoch": 0.7585658777817026,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949421076937325,
      "loss": 5.5769,
      "mean_token_accuracy": 0.18911861181259154,
      "num_tokens": 16001411.0,
      "step": 8590
    },
    {
      "entropy": 5.824386835098267,
      "epoch": 0.7590074178735429,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004949354658547753,
      "loss": 5.5408,
      "mean_token_accuracy": 0.19561953097581863,
      "num_tokens": 16009983.0,
      "step": 8595
    },
    {
      "entropy": 5.916591215133667,
      "epoch": 0.7594489579653833,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004949288197074074,
      "loss": 5.7153,
      "mean_token_accuracy": 0.17899348437786103,
      "num_tokens": 16018964.0,
      "step": 8600
    },
    {
      "entropy": 5.9272040843963625,
      "epoch": 0.7598904980572236,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949221692517589,
      "loss": 5.5511,
      "mean_token_accuracy": 0.19129472225904465,
      "num_tokens": 16027980.0,
      "step": 8605
    },
    {
      "entropy": 5.9095916748046875,
      "epoch": 0.7603320381490639,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004949155144879602,
      "loss": 5.5999,
      "mean_token_accuracy": 0.18920411467552184,
      "num_tokens": 16037271.0,
      "step": 8610
    },
    {
      "entropy": 5.903286933898926,
      "epoch": 0.7607735782409043,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949088554161417,
      "loss": 5.6872,
      "mean_token_accuracy": 0.17870717048645018,
      "num_tokens": 16046346.0,
      "step": 8615
    },
    {
      "entropy": 5.889797639846802,
      "epoch": 0.7612151183327446,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004949021920364339,
      "loss": 5.5714,
      "mean_token_accuracy": 0.19319352507591248,
      "num_tokens": 16056142.0,
      "step": 8620
    },
    {
      "entropy": 5.879462575912475,
      "epoch": 0.761656658424585,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004948955243489671,
      "loss": 5.6522,
      "mean_token_accuracy": 0.19046870321035386,
      "num_tokens": 16065487.0,
      "step": 8625
    },
    {
      "entropy": 5.948218727111817,
      "epoch": 0.7620981985164252,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948888523538721,
      "loss": 5.6862,
      "mean_token_accuracy": 0.18404763787984849,
      "num_tokens": 16075045.0,
      "step": 8630
    },
    {
      "entropy": 5.956368541717529,
      "epoch": 0.7625397386082656,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004948821760512795,
      "loss": 5.5987,
      "mean_token_accuracy": 0.18420035094022752,
      "num_tokens": 16083711.0,
      "step": 8635
    },
    {
      "entropy": 6.002153587341309,
      "epoch": 0.762981278700106,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948754954413201,
      "loss": 5.6989,
      "mean_token_accuracy": 0.18253263533115388,
      "num_tokens": 16093397.0,
      "step": 8640
    },
    {
      "entropy": 5.8907311916351315,
      "epoch": 0.7634228187919463,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004948688105241248,
      "loss": 5.6398,
      "mean_token_accuracy": 0.18502733558416368,
      "num_tokens": 16102585.0,
      "step": 8645
    },
    {
      "entropy": 5.945896291732788,
      "epoch": 0.7638643588837867,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004948621212998246,
      "loss": 5.6771,
      "mean_token_accuracy": 0.172746005654335,
      "num_tokens": 16112154.0,
      "step": 8650
    },
    {
      "entropy": 6.001024150848389,
      "epoch": 0.764305898975627,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004948554277685505,
      "loss": 5.7406,
      "mean_token_accuracy": 0.17625453025102616,
      "num_tokens": 16122371.0,
      "step": 8655
    },
    {
      "entropy": 6.0015312194824215,
      "epoch": 0.7647474390674673,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948487299304335,
      "loss": 5.7203,
      "mean_token_accuracy": 0.17225871235132217,
      "num_tokens": 16131479.0,
      "step": 8660
    },
    {
      "entropy": 5.928690242767334,
      "epoch": 0.7651889791593076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494842027785605,
      "loss": 5.7621,
      "mean_token_accuracy": 0.18196078538894653,
      "num_tokens": 16141647.0,
      "step": 8665
    },
    {
      "entropy": 5.913266897201538,
      "epoch": 0.765630519251148,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004948353213341961,
      "loss": 5.6343,
      "mean_token_accuracy": 0.18125413954257966,
      "num_tokens": 16150663.0,
      "step": 8670
    },
    {
      "entropy": 5.952857160568238,
      "epoch": 0.7660720593429884,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004948286105763385,
      "loss": 5.7484,
      "mean_token_accuracy": 0.17966294586658477,
      "num_tokens": 16160860.0,
      "step": 8675
    },
    {
      "entropy": 5.935945749282837,
      "epoch": 0.7665135994348287,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948218955121632,
      "loss": 5.5711,
      "mean_token_accuracy": 0.1831711232662201,
      "num_tokens": 16169795.0,
      "step": 8680
    },
    {
      "entropy": 5.884555101394653,
      "epoch": 0.766955139526669,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004948151761418021,
      "loss": 5.7172,
      "mean_token_accuracy": 0.18411182016134262,
      "num_tokens": 16179001.0,
      "step": 8685
    },
    {
      "entropy": 6.027784156799316,
      "epoch": 0.7673966796185093,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004948084524653868,
      "loss": 5.6437,
      "mean_token_accuracy": 0.18352477997541428,
      "num_tokens": 16188124.0,
      "step": 8690
    },
    {
      "entropy": 5.936484622955322,
      "epoch": 0.7678382197103497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004948017244830489,
      "loss": 5.6369,
      "mean_token_accuracy": 0.17818110138177873,
      "num_tokens": 16197260.0,
      "step": 8695
    },
    {
      "entropy": 5.822666025161743,
      "epoch": 0.76827975980219,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004947949921949201,
      "loss": 5.6482,
      "mean_token_accuracy": 0.18971184641122818,
      "num_tokens": 16207259.0,
      "step": 8700
    },
    {
      "entropy": 5.926111173629761,
      "epoch": 0.7687212998940304,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004947882556011325,
      "loss": 5.6817,
      "mean_token_accuracy": 0.1817387029528618,
      "num_tokens": 16216427.0,
      "step": 8705
    },
    {
      "entropy": 5.912157726287842,
      "epoch": 0.7691628399858708,
      "grad_norm": 0.890625,
      "learning_rate": 0.000494781514701818,
      "loss": 5.5838,
      "mean_token_accuracy": 0.1912463441491127,
      "num_tokens": 16225394.0,
      "step": 8710
    },
    {
      "entropy": 5.859705066680908,
      "epoch": 0.769604380077711,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947747694971085,
      "loss": 5.5786,
      "mean_token_accuracy": 0.1892262190580368,
      "num_tokens": 16234107.0,
      "step": 8715
    },
    {
      "entropy": 5.948940324783325,
      "epoch": 0.7700459201695514,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947680199871363,
      "loss": 5.7216,
      "mean_token_accuracy": 0.17314561009407042,
      "num_tokens": 16243269.0,
      "step": 8720
    },
    {
      "entropy": 5.994079875946045,
      "epoch": 0.7704874602613917,
      "grad_norm": 1.125,
      "learning_rate": 0.0004947612661720336,
      "loss": 5.6428,
      "mean_token_accuracy": 0.19023630023002625,
      "num_tokens": 16251713.0,
      "step": 8725
    },
    {
      "entropy": 5.881682014465332,
      "epoch": 0.7709290003532321,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004947545080519327,
      "loss": 5.7091,
      "mean_token_accuracy": 0.18007768392562867,
      "num_tokens": 16261630.0,
      "step": 8730
    },
    {
      "entropy": 5.98861665725708,
      "epoch": 0.7713705404450724,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000494747745626966,
      "loss": 5.696,
      "mean_token_accuracy": 0.17643645852804185,
      "num_tokens": 16271007.0,
      "step": 8735
    },
    {
      "entropy": 5.961651992797852,
      "epoch": 0.7718120805369127,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004947409788972659,
      "loss": 5.7068,
      "mean_token_accuracy": 0.17835633903741838,
      "num_tokens": 16280204.0,
      "step": 8740
    },
    {
      "entropy": 5.908645057678223,
      "epoch": 0.7722536206287531,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004947342078629649,
      "loss": 5.6641,
      "mean_token_accuracy": 0.18458299189805985,
      "num_tokens": 16289471.0,
      "step": 8745
    },
    {
      "entropy": 5.962927532196045,
      "epoch": 0.7726951607205934,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947274325241959,
      "loss": 5.6553,
      "mean_token_accuracy": 0.1837482586503029,
      "num_tokens": 16298898.0,
      "step": 8750
    },
    {
      "entropy": 5.976776361465454,
      "epoch": 0.7731367008124338,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947206528810913,
      "loss": 5.6234,
      "mean_token_accuracy": 0.18342252522706987,
      "num_tokens": 16308595.0,
      "step": 8755
    },
    {
      "entropy": 5.924045753479004,
      "epoch": 0.7735782409042741,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004947138689337843,
      "loss": 5.7095,
      "mean_token_accuracy": 0.1728400021791458,
      "num_tokens": 16317734.0,
      "step": 8760
    },
    {
      "entropy": 5.929327535629272,
      "epoch": 0.7740197809961145,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004947070806824074,
      "loss": 5.6069,
      "mean_token_accuracy": 0.18603554368019104,
      "num_tokens": 16327241.0,
      "step": 8765
    },
    {
      "entropy": 5.948109149932861,
      "epoch": 0.7744613210879547,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947002881270937,
      "loss": 5.685,
      "mean_token_accuracy": 0.18748775124549866,
      "num_tokens": 16336282.0,
      "step": 8770
    },
    {
      "entropy": 5.951261138916015,
      "epoch": 0.7749028611797951,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946934912679764,
      "loss": 5.7111,
      "mean_token_accuracy": 0.17907453179359437,
      "num_tokens": 16345494.0,
      "step": 8775
    },
    {
      "entropy": 5.92580213546753,
      "epoch": 0.7753444012716355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946866901051885,
      "loss": 5.6644,
      "mean_token_accuracy": 0.18182902932167053,
      "num_tokens": 16354796.0,
      "step": 8780
    },
    {
      "entropy": 5.876239538192749,
      "epoch": 0.7757859413634758,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946798846388634,
      "loss": 5.6211,
      "mean_token_accuracy": 0.1805059716105461,
      "num_tokens": 16364031.0,
      "step": 8785
    },
    {
      "entropy": 5.886538648605347,
      "epoch": 0.7762274814553162,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004946730748691342,
      "loss": 5.5853,
      "mean_token_accuracy": 0.18971465528011322,
      "num_tokens": 16372822.0,
      "step": 8790
    },
    {
      "entropy": 6.000423240661621,
      "epoch": 0.7766690215471564,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004946662607961344,
      "loss": 5.7166,
      "mean_token_accuracy": 0.17710226625204087,
      "num_tokens": 16382091.0,
      "step": 8795
    },
    {
      "entropy": 6.0022541046142575,
      "epoch": 0.7771105616389968,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946594424199976,
      "loss": 5.7097,
      "mean_token_accuracy": 0.17288368493318557,
      "num_tokens": 16392178.0,
      "step": 8800
    },
    {
      "entropy": 5.8729410648345945,
      "epoch": 0.7775521017308371,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004946526197408571,
      "loss": 5.5794,
      "mean_token_accuracy": 0.176253804564476,
      "num_tokens": 16401446.0,
      "step": 8805
    },
    {
      "entropy": 5.974571084976196,
      "epoch": 0.7779936418226775,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946457927588467,
      "loss": 5.5909,
      "mean_token_accuracy": 0.1833949029445648,
      "num_tokens": 16409154.0,
      "step": 8810
    },
    {
      "entropy": 5.900437641143799,
      "epoch": 0.7784351819145179,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004946389614741002,
      "loss": 5.6942,
      "mean_token_accuracy": 0.17926001846790313,
      "num_tokens": 16418507.0,
      "step": 8815
    },
    {
      "entropy": 5.937743043899536,
      "epoch": 0.7788767220063582,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004946321258867513,
      "loss": 5.6244,
      "mean_token_accuracy": 0.1828938826918602,
      "num_tokens": 16427691.0,
      "step": 8820
    },
    {
      "entropy": 5.9833784103393555,
      "epoch": 0.7793182620981985,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494625285996934,
      "loss": 5.7007,
      "mean_token_accuracy": 0.1816767781972885,
      "num_tokens": 16437248.0,
      "step": 8825
    },
    {
      "entropy": 6.008762264251709,
      "epoch": 0.7797598021900388,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946184418047823,
      "loss": 5.7185,
      "mean_token_accuracy": 0.18265429884195328,
      "num_tokens": 16445808.0,
      "step": 8830
    },
    {
      "entropy": 5.920153570175171,
      "epoch": 0.7802013422818792,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004946115933104301,
      "loss": 5.6393,
      "mean_token_accuracy": 0.18323556929826737,
      "num_tokens": 16454897.0,
      "step": 8835
    },
    {
      "entropy": 5.919833374023438,
      "epoch": 0.7806428823737195,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004946047405140119,
      "loss": 5.6643,
      "mean_token_accuracy": 0.18646166622638702,
      "num_tokens": 16464022.0,
      "step": 8840
    },
    {
      "entropy": 5.977573204040527,
      "epoch": 0.7810844224655599,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004945978834156616,
      "loss": 5.6939,
      "mean_token_accuracy": 0.17768222987651824,
      "num_tokens": 16473363.0,
      "step": 8845
    },
    {
      "entropy": 5.9358070373535154,
      "epoch": 0.7815259625574003,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004945910220155136,
      "loss": 5.7029,
      "mean_token_accuracy": 0.1819564461708069,
      "num_tokens": 16482668.0,
      "step": 8850
    },
    {
      "entropy": 5.952478122711182,
      "epoch": 0.7819675026492405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945841563137025,
      "loss": 5.6283,
      "mean_token_accuracy": 0.18354153335094453,
      "num_tokens": 16491486.0,
      "step": 8855
    },
    {
      "entropy": 5.983518743515015,
      "epoch": 0.7824090427410809,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004945772863103626,
      "loss": 5.7066,
      "mean_token_accuracy": 0.1790255442261696,
      "num_tokens": 16501132.0,
      "step": 8860
    },
    {
      "entropy": 5.859392881393433,
      "epoch": 0.7828505828329212,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945704120056286,
      "loss": 5.5763,
      "mean_token_accuracy": 0.19056922793388367,
      "num_tokens": 16510717.0,
      "step": 8865
    },
    {
      "entropy": 5.99395170211792,
      "epoch": 0.7832921229247616,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494563533399635,
      "loss": 5.6973,
      "mean_token_accuracy": 0.18568243086338043,
      "num_tokens": 16519314.0,
      "step": 8870
    },
    {
      "entropy": 5.945958375930786,
      "epoch": 0.7837336630166019,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004945566504925167,
      "loss": 5.6923,
      "mean_token_accuracy": 0.18825736194849013,
      "num_tokens": 16528304.0,
      "step": 8875
    },
    {
      "entropy": 5.896948528289795,
      "epoch": 0.7841752031084422,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945497632844084,
      "loss": 5.6105,
      "mean_token_accuracy": 0.1896432802081108,
      "num_tokens": 16537312.0,
      "step": 8880
    },
    {
      "entropy": 5.9886924743652346,
      "epoch": 0.7846167432002826,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945428717754453,
      "loss": 5.7037,
      "mean_token_accuracy": 0.18342699855566025,
      "num_tokens": 16546886.0,
      "step": 8885
    },
    {
      "entropy": 5.915195751190185,
      "epoch": 0.7850582832921229,
      "grad_norm": 0.96875,
      "learning_rate": 0.000494535975965762,
      "loss": 5.6037,
      "mean_token_accuracy": 0.18885530531406403,
      "num_tokens": 16555556.0,
      "step": 8890
    },
    {
      "entropy": 5.863842582702636,
      "epoch": 0.7854998233839633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945290758554939,
      "loss": 5.6618,
      "mean_token_accuracy": 0.18245059251785278,
      "num_tokens": 16564030.0,
      "step": 8895
    },
    {
      "entropy": 5.936520957946778,
      "epoch": 0.7859413634758036,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494522171444776,
      "loss": 5.5898,
      "mean_token_accuracy": 0.184000164270401,
      "num_tokens": 16573775.0,
      "step": 8900
    },
    {
      "entropy": 5.878285503387451,
      "epoch": 0.786382903567644,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004945152627337437,
      "loss": 5.654,
      "mean_token_accuracy": 0.18854805678129197,
      "num_tokens": 16583946.0,
      "step": 8905
    },
    {
      "entropy": 5.931978797912597,
      "epoch": 0.7868244436594842,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004945083497225321,
      "loss": 5.6531,
      "mean_token_accuracy": 0.17515482604503632,
      "num_tokens": 16592263.0,
      "step": 8910
    },
    {
      "entropy": 5.96363091468811,
      "epoch": 0.7872659837513246,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945014324112769,
      "loss": 5.6195,
      "mean_token_accuracy": 0.1868958055973053,
      "num_tokens": 16601690.0,
      "step": 8915
    },
    {
      "entropy": 5.90001072883606,
      "epoch": 0.787707523843165,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004944945108001134,
      "loss": 5.6727,
      "mean_token_accuracy": 0.17649522721767424,
      "num_tokens": 16611211.0,
      "step": 8920
    },
    {
      "entropy": 5.9767755508422855,
      "epoch": 0.7881490639350053,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944875848891773,
      "loss": 5.7061,
      "mean_token_accuracy": 0.17708539515733718,
      "num_tokens": 16620477.0,
      "step": 8925
    },
    {
      "entropy": 6.068809175491333,
      "epoch": 0.7885906040268457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944806546786042,
      "loss": 5.7332,
      "mean_token_accuracy": 0.1714392051100731,
      "num_tokens": 16629332.0,
      "step": 8930
    },
    {
      "entropy": 5.922574234008789,
      "epoch": 0.7890321441186859,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049447372016853,
      "loss": 5.7174,
      "mean_token_accuracy": 0.18329513370990752,
      "num_tokens": 16639512.0,
      "step": 8935
    },
    {
      "entropy": 5.923654651641845,
      "epoch": 0.7894736842105263,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944667813590904,
      "loss": 5.7306,
      "mean_token_accuracy": 0.17635459154844285,
      "num_tokens": 16648352.0,
      "step": 8940
    },
    {
      "entropy": 5.974083948135376,
      "epoch": 0.7899152243023666,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944598382504214,
      "loss": 5.5571,
      "mean_token_accuracy": 0.19209358096122742,
      "num_tokens": 16657613.0,
      "step": 8945
    },
    {
      "entropy": 5.936112976074218,
      "epoch": 0.790356764394207,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944528908426589,
      "loss": 5.6611,
      "mean_token_accuracy": 0.1815002366900444,
      "num_tokens": 16666673.0,
      "step": 8950
    },
    {
      "entropy": 5.935863590240478,
      "epoch": 0.7907983044860474,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944459391359391,
      "loss": 5.6741,
      "mean_token_accuracy": 0.1859077423810959,
      "num_tokens": 16676244.0,
      "step": 8955
    },
    {
      "entropy": 6.038500070571899,
      "epoch": 0.7912398445778877,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004944389831303982,
      "loss": 5.8081,
      "mean_token_accuracy": 0.17108060270547867,
      "num_tokens": 16685393.0,
      "step": 8960
    },
    {
      "entropy": 5.980288410186768,
      "epoch": 0.791681384669728,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004944320228261723,
      "loss": 5.7108,
      "mean_token_accuracy": 0.180125792324543,
      "num_tokens": 16695597.0,
      "step": 8965
    },
    {
      "entropy": 5.957344102859497,
      "epoch": 0.7921229247615683,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000494425058223398,
      "loss": 5.646,
      "mean_token_accuracy": 0.1851449728012085,
      "num_tokens": 16705473.0,
      "step": 8970
    },
    {
      "entropy": 5.930376482009888,
      "epoch": 0.7925644648534087,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944180893222117,
      "loss": 5.6917,
      "mean_token_accuracy": 0.18281938135623932,
      "num_tokens": 16714913.0,
      "step": 8975
    },
    {
      "entropy": 5.960370588302612,
      "epoch": 0.793006004945249,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944111161227496,
      "loss": 5.6073,
      "mean_token_accuracy": 0.18728964924812316,
      "num_tokens": 16724993.0,
      "step": 8980
    },
    {
      "entropy": 5.929873323440551,
      "epoch": 0.7934475450370894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944041386251486,
      "loss": 5.6538,
      "mean_token_accuracy": 0.18389132022857665,
      "num_tokens": 16734077.0,
      "step": 8985
    },
    {
      "entropy": 5.847494316101074,
      "epoch": 0.7938890851289298,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943971568295454,
      "loss": 5.5649,
      "mean_token_accuracy": 0.19323885887861253,
      "num_tokens": 16743936.0,
      "step": 8990
    },
    {
      "entropy": 5.89783616065979,
      "epoch": 0.79433062522077,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004943901707360765,
      "loss": 5.5146,
      "mean_token_accuracy": 0.19727666079998016,
      "num_tokens": 16752718.0,
      "step": 8995
    },
    {
      "entropy": 5.930490970611572,
      "epoch": 0.7947721653126104,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494383180344879,
      "loss": 5.6294,
      "mean_token_accuracy": 0.19191263020038604,
      "num_tokens": 16762162.0,
      "step": 9000
    },
    {
      "epoch": 0.7947721653126104,
      "eval_entropy": 5.717408272958153,
      "eval_loss": 5.655894756317139,
      "eval_mean_token_accuracy": 0.1915602565562671,
      "eval_num_tokens": 16762162.0,
      "eval_runtime": 20.283,
      "eval_samples_per_second": 1741.117,
      "eval_steps_per_second": 217.67,
      "step": 9000
    },
    {
      "entropy": 5.871576881408691,
      "epoch": 0.7952137054044507,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004943761856560898,
      "loss": 5.6329,
      "mean_token_accuracy": 0.19682139307260513,
      "num_tokens": 16770722.0,
      "step": 9005
    },
    {
      "entropy": 5.9635063171386715,
      "epoch": 0.7956552454962911,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000494369186669846,
      "loss": 5.6934,
      "mean_token_accuracy": 0.18363819718360902,
      "num_tokens": 16779772.0,
      "step": 9010
    },
    {
      "entropy": 5.954889583587646,
      "epoch": 0.7960967855881314,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943621833862844,
      "loss": 5.6714,
      "mean_token_accuracy": 0.18223804533481597,
      "num_tokens": 16789764.0,
      "step": 9015
    },
    {
      "entropy": 5.912001276016236,
      "epoch": 0.7965383256799717,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004943551758055425,
      "loss": 5.5998,
      "mean_token_accuracy": 0.18742647171020507,
      "num_tokens": 16798055.0,
      "step": 9020
    },
    {
      "entropy": 5.829352331161499,
      "epoch": 0.7969798657718121,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004943481639277574,
      "loss": 5.5809,
      "mean_token_accuracy": 0.1951695576310158,
      "num_tokens": 16807750.0,
      "step": 9025
    },
    {
      "entropy": 5.8902161598205565,
      "epoch": 0.7974214058636524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004943411477530665,
      "loss": 5.5036,
      "mean_token_accuracy": 0.20224213600158691,
      "num_tokens": 16816512.0,
      "step": 9030
    },
    {
      "entropy": 5.940272569656372,
      "epoch": 0.7978629459554928,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943341272816073,
      "loss": 5.617,
      "mean_token_accuracy": 0.18778126686811447,
      "num_tokens": 16825875.0,
      "step": 9035
    },
    {
      "entropy": 5.955309629440308,
      "epoch": 0.7983044860473331,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943271025135173,
      "loss": 5.6204,
      "mean_token_accuracy": 0.18766249418258668,
      "num_tokens": 16836395.0,
      "step": 9040
    },
    {
      "entropy": 5.941666841506958,
      "epoch": 0.7987460261391734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943200734489341,
      "loss": 5.5818,
      "mean_token_accuracy": 0.1908138021826744,
      "num_tokens": 16844505.0,
      "step": 9045
    },
    {
      "entropy": 5.868502902984619,
      "epoch": 0.7991875662310137,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004943130400879953,
      "loss": 5.6897,
      "mean_token_accuracy": 0.1812623083591461,
      "num_tokens": 16854828.0,
      "step": 9050
    },
    {
      "entropy": 5.898808431625366,
      "epoch": 0.7996291063228541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943060024308389,
      "loss": 5.668,
      "mean_token_accuracy": 0.18504270762205124,
      "num_tokens": 16863596.0,
      "step": 9055
    },
    {
      "entropy": 5.928490352630615,
      "epoch": 0.8000706464146945,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004942989604776026,
      "loss": 5.624,
      "mean_token_accuracy": 0.186293625831604,
      "num_tokens": 16872726.0,
      "step": 9060
    },
    {
      "entropy": 5.917617607116699,
      "epoch": 0.8005121865065348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004942919142284244,
      "loss": 5.6129,
      "mean_token_accuracy": 0.18066419959068297,
      "num_tokens": 16880916.0,
      "step": 9065
    },
    {
      "entropy": 5.982567930221558,
      "epoch": 0.8009537265983752,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942848636834423,
      "loss": 5.6601,
      "mean_token_accuracy": 0.1804071068763733,
      "num_tokens": 16890051.0,
      "step": 9070
    },
    {
      "entropy": 5.9492415428161625,
      "epoch": 0.8013952666902154,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942778088427944,
      "loss": 5.7529,
      "mean_token_accuracy": 0.18031782805919647,
      "num_tokens": 16899836.0,
      "step": 9075
    },
    {
      "entropy": 5.963057613372802,
      "epoch": 0.8018368067820558,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494270749706619,
      "loss": 5.6435,
      "mean_token_accuracy": 0.18878489285707473,
      "num_tokens": 16908114.0,
      "step": 9080
    },
    {
      "entropy": 5.930170202255249,
      "epoch": 0.8022783468738961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004942636862750544,
      "loss": 5.6736,
      "mean_token_accuracy": 0.18289379477500917,
      "num_tokens": 16916863.0,
      "step": 9085
    },
    {
      "entropy": 5.916874122619629,
      "epoch": 0.8027198869657365,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004942566185482387,
      "loss": 5.6126,
      "mean_token_accuracy": 0.1851662814617157,
      "num_tokens": 16926397.0,
      "step": 9090
    },
    {
      "entropy": 5.927210807800293,
      "epoch": 0.8031614270575769,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942495465263106,
      "loss": 5.6441,
      "mean_token_accuracy": 0.18736824989318848,
      "num_tokens": 16936326.0,
      "step": 9095
    },
    {
      "entropy": 5.985590505599975,
      "epoch": 0.8036029671494171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942424702094085,
      "loss": 5.6676,
      "mean_token_accuracy": 0.18307853490114212,
      "num_tokens": 16945476.0,
      "step": 9100
    },
    {
      "entropy": 5.902594041824341,
      "epoch": 0.8040445072412575,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942353895976713,
      "loss": 5.6149,
      "mean_token_accuracy": 0.1900067999958992,
      "num_tokens": 16955006.0,
      "step": 9105
    },
    {
      "entropy": 5.999951791763306,
      "epoch": 0.8044860473330978,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942283046912374,
      "loss": 5.6614,
      "mean_token_accuracy": 0.18079947382211686,
      "num_tokens": 16963609.0,
      "step": 9110
    },
    {
      "entropy": 5.984429883956909,
      "epoch": 0.8049275874249382,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004942212154902456,
      "loss": 5.6698,
      "mean_token_accuracy": 0.18488241136074066,
      "num_tokens": 16973227.0,
      "step": 9115
    },
    {
      "entropy": 5.8114728927612305,
      "epoch": 0.8053691275167785,
      "grad_norm": 0.984375,
      "learning_rate": 0.000494214121994835,
      "loss": 5.6181,
      "mean_token_accuracy": 0.18809541165828705,
      "num_tokens": 16983429.0,
      "step": 9120
    },
    {
      "entropy": 5.980748224258423,
      "epoch": 0.8058106676086189,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942070242051444,
      "loss": 5.6883,
      "mean_token_accuracy": 0.1820809170603752,
      "num_tokens": 16993395.0,
      "step": 9125
    },
    {
      "entropy": 5.988777780532837,
      "epoch": 0.8062522077004592,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004941999221213129,
      "loss": 5.7221,
      "mean_token_accuracy": 0.17346379458904265,
      "num_tokens": 17004192.0,
      "step": 9130
    },
    {
      "entropy": 5.90602216720581,
      "epoch": 0.8066937477922995,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004941928157434796,
      "loss": 5.6329,
      "mean_token_accuracy": 0.1836729899048805,
      "num_tokens": 17012976.0,
      "step": 9135
    },
    {
      "entropy": 5.877796602249146,
      "epoch": 0.8071352878841399,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004941857050717836,
      "loss": 5.6001,
      "mean_token_accuracy": 0.18954011797904968,
      "num_tokens": 17022284.0,
      "step": 9140
    },
    {
      "entropy": 5.880238199234009,
      "epoch": 0.8075768279759802,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004941785901063642,
      "loss": 5.5673,
      "mean_token_accuracy": 0.18442789018154143,
      "num_tokens": 17030948.0,
      "step": 9145
    },
    {
      "entropy": 5.9989025592803955,
      "epoch": 0.8080183680678206,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004941714708473611,
      "loss": 5.7629,
      "mean_token_accuracy": 0.17965551763772963,
      "num_tokens": 17039685.0,
      "step": 9150
    },
    {
      "entropy": 5.931429624557495,
      "epoch": 0.8084599081596608,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941643472949134,
      "loss": 5.591,
      "mean_token_accuracy": 0.19194363057613373,
      "num_tokens": 17049048.0,
      "step": 9155
    },
    {
      "entropy": 5.880387020111084,
      "epoch": 0.8089014482515012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941572194491608,
      "loss": 5.6225,
      "mean_token_accuracy": 0.18224881291389466,
      "num_tokens": 17057719.0,
      "step": 9160
    },
    {
      "entropy": 5.947750282287598,
      "epoch": 0.8093429883433416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000494150087310243,
      "loss": 5.6224,
      "mean_token_accuracy": 0.18389955908060074,
      "num_tokens": 17066956.0,
      "step": 9165
    },
    {
      "entropy": 5.973948383331299,
      "epoch": 0.8097845284351819,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941429508782995,
      "loss": 5.6501,
      "mean_token_accuracy": 0.18145118802785873,
      "num_tokens": 17076165.0,
      "step": 9170
    },
    {
      "entropy": 5.948069524765015,
      "epoch": 0.8102260685270223,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941358101534703,
      "loss": 5.679,
      "mean_token_accuracy": 0.17935617566108703,
      "num_tokens": 17086000.0,
      "step": 9175
    },
    {
      "entropy": 5.989986228942871,
      "epoch": 0.8106676086188626,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941286651358952,
      "loss": 5.6682,
      "mean_token_accuracy": 0.17963472306728362,
      "num_tokens": 17095496.0,
      "step": 9180
    },
    {
      "entropy": 5.890500593185425,
      "epoch": 0.8111091487107029,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004941215158257142,
      "loss": 5.5638,
      "mean_token_accuracy": 0.19024398624897004,
      "num_tokens": 17104590.0,
      "step": 9185
    },
    {
      "entropy": 5.915885162353516,
      "epoch": 0.8115506888025432,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941143622230672,
      "loss": 5.6618,
      "mean_token_accuracy": 0.18562659472227097,
      "num_tokens": 17113678.0,
      "step": 9190
    },
    {
      "entropy": 5.993251323699951,
      "epoch": 0.8119922288943836,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941072043280946,
      "loss": 5.7286,
      "mean_token_accuracy": 0.17823514044284822,
      "num_tokens": 17123826.0,
      "step": 9195
    },
    {
      "entropy": 5.935890245437622,
      "epoch": 0.812433768986224,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004941000421409365,
      "loss": 5.6349,
      "mean_token_accuracy": 0.18861936330795287,
      "num_tokens": 17133151.0,
      "step": 9200
    },
    {
      "entropy": 5.903849840164185,
      "epoch": 0.8128753090780643,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004940928756617331,
      "loss": 5.6691,
      "mean_token_accuracy": 0.18083446621894836,
      "num_tokens": 17143265.0,
      "step": 9205
    },
    {
      "entropy": 5.9282097816467285,
      "epoch": 0.8133168491699047,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940857048906248,
      "loss": 5.6228,
      "mean_token_accuracy": 0.18259122222661972,
      "num_tokens": 17152817.0,
      "step": 9210
    },
    {
      "entropy": 5.928669595718384,
      "epoch": 0.8137583892617449,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004940785298277523,
      "loss": 5.6884,
      "mean_token_accuracy": 0.17689198553562163,
      "num_tokens": 17161906.0,
      "step": 9215
    },
    {
      "entropy": 5.910034418106079,
      "epoch": 0.8141999293535853,
      "grad_norm": 1.0,
      "learning_rate": 0.000494071350473256,
      "loss": 5.6461,
      "mean_token_accuracy": 0.181121264398098,
      "num_tokens": 17171341.0,
      "step": 9220
    },
    {
      "entropy": 5.979455280303955,
      "epoch": 0.8146414694454256,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004940641668272765,
      "loss": 5.6771,
      "mean_token_accuracy": 0.18112974613904953,
      "num_tokens": 17181498.0,
      "step": 9225
    },
    {
      "entropy": 5.887812089920044,
      "epoch": 0.815083009537266,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940569788899544,
      "loss": 5.541,
      "mean_token_accuracy": 0.19354602545499802,
      "num_tokens": 17191551.0,
      "step": 9230
    },
    {
      "entropy": 5.936913776397705,
      "epoch": 0.8155245496291064,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940497866614309,
      "loss": 5.6735,
      "mean_token_accuracy": 0.18211649656295775,
      "num_tokens": 17201192.0,
      "step": 9235
    },
    {
      "entropy": 5.821828699111938,
      "epoch": 0.8159660897209466,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940425901418465,
      "loss": 5.56,
      "mean_token_accuracy": 0.19922360628843308,
      "num_tokens": 17210003.0,
      "step": 9240
    },
    {
      "entropy": 5.88896918296814,
      "epoch": 0.816407629812787,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940353893313423,
      "loss": 5.6283,
      "mean_token_accuracy": 0.18848212510347367,
      "num_tokens": 17219616.0,
      "step": 9245
    },
    {
      "entropy": 5.894268321990967,
      "epoch": 0.8168491699046273,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004940281842300596,
      "loss": 5.4819,
      "mean_token_accuracy": 0.19770817458629608,
      "num_tokens": 17228465.0,
      "step": 9250
    },
    {
      "entropy": 5.9158226490020756,
      "epoch": 0.8172907099964677,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004940209748381392,
      "loss": 5.6743,
      "mean_token_accuracy": 0.17920099943876266,
      "num_tokens": 17237673.0,
      "step": 9255
    },
    {
      "entropy": 5.854257822036743,
      "epoch": 0.817732250088308,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004940137611557225,
      "loss": 5.5452,
      "mean_token_accuracy": 0.19255712926387786,
      "num_tokens": 17246286.0,
      "step": 9260
    },
    {
      "entropy": 5.924583911895752,
      "epoch": 0.8181737901801484,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004940065431829508,
      "loss": 5.6604,
      "mean_token_accuracy": 0.1844808742403984,
      "num_tokens": 17256193.0,
      "step": 9265
    },
    {
      "entropy": 5.969257354736328,
      "epoch": 0.8186153302719887,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004939993209199654,
      "loss": 5.6382,
      "mean_token_accuracy": 0.1871926099061966,
      "num_tokens": 17265006.0,
      "step": 9270
    },
    {
      "entropy": 5.941160583496094,
      "epoch": 0.819056870363829,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004939920943669079,
      "loss": 5.6397,
      "mean_token_accuracy": 0.18583282679319382,
      "num_tokens": 17274173.0,
      "step": 9275
    },
    {
      "entropy": 5.855961799621582,
      "epoch": 0.8194984104556694,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939848635239198,
      "loss": 5.7061,
      "mean_token_accuracy": 0.17912741005420685,
      "num_tokens": 17284024.0,
      "step": 9280
    },
    {
      "entropy": 5.892509984970093,
      "epoch": 0.8199399505475097,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004939776283911429,
      "loss": 5.636,
      "mean_token_accuracy": 0.18788398653268815,
      "num_tokens": 17294096.0,
      "step": 9285
    },
    {
      "entropy": 6.0269073963165285,
      "epoch": 0.8203814906393501,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004939703889687186,
      "loss": 5.6807,
      "mean_token_accuracy": 0.17701893597841262,
      "num_tokens": 17303740.0,
      "step": 9290
    },
    {
      "entropy": 5.969491958618164,
      "epoch": 0.8208230307311903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004939631452567891,
      "loss": 5.6562,
      "mean_token_accuracy": 0.17997532486915588,
      "num_tokens": 17313276.0,
      "step": 9295
    },
    {
      "entropy": 5.8992119312286375,
      "epoch": 0.8212645708230307,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493955897255496,
      "loss": 5.6213,
      "mean_token_accuracy": 0.18585679531097413,
      "num_tokens": 17322008.0,
      "step": 9300
    },
    {
      "entropy": 5.819156980514526,
      "epoch": 0.8217061109148711,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939486449649814,
      "loss": 5.5435,
      "mean_token_accuracy": 0.19549853801727296,
      "num_tokens": 17331753.0,
      "step": 9305
    },
    {
      "entropy": 5.97447361946106,
      "epoch": 0.8221476510067114,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004939413883853873,
      "loss": 5.6908,
      "mean_token_accuracy": 0.17804472297430038,
      "num_tokens": 17340424.0,
      "step": 9310
    },
    {
      "entropy": 5.970905017852783,
      "epoch": 0.8225891910985518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493934127516856,
      "loss": 5.6909,
      "mean_token_accuracy": 0.18500348627567292,
      "num_tokens": 17350407.0,
      "step": 9315
    },
    {
      "entropy": 5.98191089630127,
      "epoch": 0.823030731190392,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939268623595297,
      "loss": 5.7289,
      "mean_token_accuracy": 0.1750431999564171,
      "num_tokens": 17360729.0,
      "step": 9320
    },
    {
      "entropy": 5.990963554382324,
      "epoch": 0.8234722712822324,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939195929135506,
      "loss": 5.6494,
      "mean_token_accuracy": 0.17734821289777755,
      "num_tokens": 17369338.0,
      "step": 9325
    },
    {
      "entropy": 5.976259994506836,
      "epoch": 0.8239138113740727,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004939123191790612,
      "loss": 5.6738,
      "mean_token_accuracy": 0.18052253872156143,
      "num_tokens": 17379347.0,
      "step": 9330
    },
    {
      "entropy": 5.941456604003906,
      "epoch": 0.8243553514659131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939050411562041,
      "loss": 5.5381,
      "mean_token_accuracy": 0.19379150122404099,
      "num_tokens": 17387483.0,
      "step": 9335
    },
    {
      "entropy": 5.8574529647827145,
      "epoch": 0.8247968915577535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004938977588451216,
      "loss": 5.6197,
      "mean_token_accuracy": 0.1892271786928177,
      "num_tokens": 17395635.0,
      "step": 9340
    },
    {
      "entropy": 5.939624404907226,
      "epoch": 0.8252384316495938,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004938904722459565,
      "loss": 5.653,
      "mean_token_accuracy": 0.18270640522241594,
      "num_tokens": 17404363.0,
      "step": 9345
    },
    {
      "entropy": 6.022149848937988,
      "epoch": 0.8256799717414341,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004938831813588515,
      "loss": 5.6425,
      "mean_token_accuracy": 0.1869293823838234,
      "num_tokens": 17413448.0,
      "step": 9350
    },
    {
      "entropy": 5.8765403747558596,
      "epoch": 0.8261215118332744,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938758861839495,
      "loss": 5.5984,
      "mean_token_accuracy": 0.18677931874990464,
      "num_tokens": 17423408.0,
      "step": 9355
    },
    {
      "entropy": 5.9144378185272215,
      "epoch": 0.8265630519251148,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938685867213934,
      "loss": 5.7293,
      "mean_token_accuracy": 0.17140262573957443,
      "num_tokens": 17434449.0,
      "step": 9360
    },
    {
      "entropy": 6.008839702606201,
      "epoch": 0.8270045920169551,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004938612829713262,
      "loss": 5.5745,
      "mean_token_accuracy": 0.18577081710100174,
      "num_tokens": 17443338.0,
      "step": 9365
    },
    {
      "entropy": 5.913425159454346,
      "epoch": 0.8274461321087955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938539749338909,
      "loss": 5.681,
      "mean_token_accuracy": 0.1822756841778755,
      "num_tokens": 17453385.0,
      "step": 9370
    },
    {
      "entropy": 5.9598291397094725,
      "epoch": 0.8278876722006359,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938466626092308,
      "loss": 5.6298,
      "mean_token_accuracy": 0.18665705770254135,
      "num_tokens": 17462167.0,
      "step": 9375
    },
    {
      "entropy": 5.898651742935181,
      "epoch": 0.8283292122924761,
      "grad_norm": 0.875,
      "learning_rate": 0.000493839345997489,
      "loss": 5.5839,
      "mean_token_accuracy": 0.18590615689754486,
      "num_tokens": 17471572.0,
      "step": 9380
    },
    {
      "entropy": 5.881286191940307,
      "epoch": 0.8287707523843165,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004938320250988087,
      "loss": 5.5873,
      "mean_token_accuracy": 0.19194208532571794,
      "num_tokens": 17480743.0,
      "step": 9385
    },
    {
      "entropy": 5.976055049896241,
      "epoch": 0.8292122924761568,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004938246999133337,
      "loss": 5.7161,
      "mean_token_accuracy": 0.17756921499967576,
      "num_tokens": 17491423.0,
      "step": 9390
    },
    {
      "entropy": 5.964974641799927,
      "epoch": 0.8296538325679972,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004938173704412071,
      "loss": 5.6525,
      "mean_token_accuracy": 0.1818349227309227,
      "num_tokens": 17501361.0,
      "step": 9395
    },
    {
      "entropy": 5.887234020233154,
      "epoch": 0.8300953726598375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004938100366825728,
      "loss": 5.6098,
      "mean_token_accuracy": 0.19193477630615235,
      "num_tokens": 17511078.0,
      "step": 9400
    },
    {
      "entropy": 5.92705979347229,
      "epoch": 0.8305369127516778,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938026986375742,
      "loss": 5.6391,
      "mean_token_accuracy": 0.1834984079003334,
      "num_tokens": 17520571.0,
      "step": 9405
    },
    {
      "entropy": 5.978323459625244,
      "epoch": 0.8309784528435182,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937953563063553,
      "loss": 5.693,
      "mean_token_accuracy": 0.17575069814920424,
      "num_tokens": 17530420.0,
      "step": 9410
    },
    {
      "entropy": 5.909533071517944,
      "epoch": 0.8314199929353585,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004937880096890598,
      "loss": 5.6011,
      "mean_token_accuracy": 0.19071900397539138,
      "num_tokens": 17539635.0,
      "step": 9415
    },
    {
      "entropy": 6.007608079910279,
      "epoch": 0.8318615330271989,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937806587858315,
      "loss": 5.7047,
      "mean_token_accuracy": 0.17142973244190216,
      "num_tokens": 17548703.0,
      "step": 9420
    },
    {
      "entropy": 5.905121231079102,
      "epoch": 0.8323030731190392,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004937733035968147,
      "loss": 5.6516,
      "mean_token_accuracy": 0.1837680846452713,
      "num_tokens": 17558319.0,
      "step": 9425
    },
    {
      "entropy": 5.911935186386108,
      "epoch": 0.8327446132108796,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004937659441221531,
      "loss": 5.5783,
      "mean_token_accuracy": 0.19150104224681855,
      "num_tokens": 17566932.0,
      "step": 9430
    },
    {
      "entropy": 5.95340051651001,
      "epoch": 0.8331861533027198,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004937585803619912,
      "loss": 5.6643,
      "mean_token_accuracy": 0.18719276636838914,
      "num_tokens": 17577126.0,
      "step": 9435
    },
    {
      "entropy": 5.919023180007935,
      "epoch": 0.8336276933945602,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004937512123164731,
      "loss": 5.6104,
      "mean_token_accuracy": 0.1872269779443741,
      "num_tokens": 17586016.0,
      "step": 9440
    },
    {
      "entropy": 5.946781635284424,
      "epoch": 0.8340692334864006,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004937438399857431,
      "loss": 5.7584,
      "mean_token_accuracy": 0.17964789569377898,
      "num_tokens": 17595042.0,
      "step": 9445
    },
    {
      "entropy": 5.896650505065918,
      "epoch": 0.8345107735782409,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004937364633699459,
      "loss": 5.5339,
      "mean_token_accuracy": 0.18852169215679168,
      "num_tokens": 17604808.0,
      "step": 9450
    },
    {
      "entropy": 5.9106419563293455,
      "epoch": 0.8349523136700813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004937290824692255,
      "loss": 5.6464,
      "mean_token_accuracy": 0.18648144602775574,
      "num_tokens": 17613937.0,
      "step": 9455
    },
    {
      "entropy": 6.005748510360718,
      "epoch": 0.8353938537619215,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937216972837269,
      "loss": 5.687,
      "mean_token_accuracy": 0.1841978460550308,
      "num_tokens": 17623572.0,
      "step": 9460
    },
    {
      "entropy": 5.897420501708984,
      "epoch": 0.8358353938537619,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937143078135946,
      "loss": 5.577,
      "mean_token_accuracy": 0.19123802930116654,
      "num_tokens": 17633371.0,
      "step": 9465
    },
    {
      "entropy": 5.942810153961181,
      "epoch": 0.8362769339456022,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937069140589735,
      "loss": 5.6859,
      "mean_token_accuracy": 0.17899446785449982,
      "num_tokens": 17642481.0,
      "step": 9470
    },
    {
      "entropy": 5.938759803771973,
      "epoch": 0.8367184740374426,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004936995160200083,
      "loss": 5.6528,
      "mean_token_accuracy": 0.17990794777870178,
      "num_tokens": 17651620.0,
      "step": 9475
    },
    {
      "entropy": 5.906102800369263,
      "epoch": 0.837160014129283,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004936921136968439,
      "loss": 5.57,
      "mean_token_accuracy": 0.19319702088832855,
      "num_tokens": 17661076.0,
      "step": 9480
    },
    {
      "entropy": 5.927499437332154,
      "epoch": 0.8376015542211233,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004936847070896254,
      "loss": 5.6265,
      "mean_token_accuracy": 0.18661921620368957,
      "num_tokens": 17670087.0,
      "step": 9485
    },
    {
      "entropy": 5.971121120452881,
      "epoch": 0.8380430943129636,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004936772961984979,
      "loss": 5.7637,
      "mean_token_accuracy": 0.1712151512503624,
      "num_tokens": 17679931.0,
      "step": 9490
    },
    {
      "entropy": 5.942320108413696,
      "epoch": 0.8384846344048039,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936698810236065,
      "loss": 5.6331,
      "mean_token_accuracy": 0.18381179720163346,
      "num_tokens": 17688185.0,
      "step": 9495
    },
    {
      "entropy": 5.912482357025146,
      "epoch": 0.8389261744966443,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936624615650964,
      "loss": 5.5743,
      "mean_token_accuracy": 0.1883551299571991,
      "num_tokens": 17697245.0,
      "step": 9500
    },
    {
      "entropy": 5.938527345657349,
      "epoch": 0.8393677145884846,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004936550378231133,
      "loss": 5.6426,
      "mean_token_accuracy": 0.18364092707633972,
      "num_tokens": 17706348.0,
      "step": 9505
    },
    {
      "entropy": 5.860145187377929,
      "epoch": 0.839809254680325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004936476097978023,
      "loss": 5.5914,
      "mean_token_accuracy": 0.1840891122817993,
      "num_tokens": 17715139.0,
      "step": 9510
    },
    {
      "entropy": 5.979407882690429,
      "epoch": 0.8402507947721654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936401774893088,
      "loss": 5.6625,
      "mean_token_accuracy": 0.1891412079334259,
      "num_tokens": 17723944.0,
      "step": 9515
    },
    {
      "entropy": 5.938834762573242,
      "epoch": 0.8406923348640056,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936327408977786,
      "loss": 5.6815,
      "mean_token_accuracy": 0.1792125716805458,
      "num_tokens": 17733298.0,
      "step": 9520
    },
    {
      "entropy": 5.988562107086182,
      "epoch": 0.841133874955846,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936253000233574,
      "loss": 5.7052,
      "mean_token_accuracy": 0.1826997458934784,
      "num_tokens": 17743501.0,
      "step": 9525
    },
    {
      "entropy": 5.9449090480804445,
      "epoch": 0.8415754150476863,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004936178548661909,
      "loss": 5.5581,
      "mean_token_accuracy": 0.19213056862354277,
      "num_tokens": 17752071.0,
      "step": 9530
    },
    {
      "entropy": 5.929076766967773,
      "epoch": 0.8420169551395267,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493610405426425,
      "loss": 5.6172,
      "mean_token_accuracy": 0.19263315349817275,
      "num_tokens": 17760754.0,
      "step": 9535
    },
    {
      "entropy": 5.928290605545044,
      "epoch": 0.842458495231367,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004936029517042056,
      "loss": 5.686,
      "mean_token_accuracy": 0.18308390378952027,
      "num_tokens": 17770940.0,
      "step": 9540
    },
    {
      "entropy": 5.930697727203369,
      "epoch": 0.8429000353232073,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004935954936996787,
      "loss": 5.6458,
      "mean_token_accuracy": 0.18414505422115326,
      "num_tokens": 17780197.0,
      "step": 9545
    },
    {
      "entropy": 5.878594207763672,
      "epoch": 0.8433415754150477,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935880314129903,
      "loss": 5.5952,
      "mean_token_accuracy": 0.18871242552995682,
      "num_tokens": 17789960.0,
      "step": 9550
    },
    {
      "entropy": 5.941827869415283,
      "epoch": 0.843783115506888,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935805648442868,
      "loss": 5.5795,
      "mean_token_accuracy": 0.1938926547765732,
      "num_tokens": 17799364.0,
      "step": 9555
    },
    {
      "entropy": 5.972416353225708,
      "epoch": 0.8442246555987284,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004935730939937143,
      "loss": 5.6432,
      "mean_token_accuracy": 0.18220338225364685,
      "num_tokens": 17808359.0,
      "step": 9560
    },
    {
      "entropy": 5.92724986076355,
      "epoch": 0.8446661956905687,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935656188614192,
      "loss": 5.7196,
      "mean_token_accuracy": 0.17525357007980347,
      "num_tokens": 17818098.0,
      "step": 9565
    },
    {
      "entropy": 5.961409091949463,
      "epoch": 0.845107735782409,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004935581394475479,
      "loss": 5.5942,
      "mean_token_accuracy": 0.1823800802230835,
      "num_tokens": 17827443.0,
      "step": 9570
    },
    {
      "entropy": 5.925687551498413,
      "epoch": 0.8455492758742493,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935506557522469,
      "loss": 5.6061,
      "mean_token_accuracy": 0.1885109946131706,
      "num_tokens": 17836683.0,
      "step": 9575
    },
    {
      "entropy": 5.915205192565918,
      "epoch": 0.8459908159660897,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493543167775663,
      "loss": 5.7199,
      "mean_token_accuracy": 0.17950170785188674,
      "num_tokens": 17845949.0,
      "step": 9580
    },
    {
      "entropy": 5.896881866455078,
      "epoch": 0.8464323560579301,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004935356755179426,
      "loss": 5.634,
      "mean_token_accuracy": 0.1836997836828232,
      "num_tokens": 17855309.0,
      "step": 9585
    },
    {
      "entropy": 5.951823997497558,
      "epoch": 0.8468738961497704,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004935281789792326,
      "loss": 5.5803,
      "mean_token_accuracy": 0.1880348429083824,
      "num_tokens": 17865455.0,
      "step": 9590
    },
    {
      "entropy": 5.90617938041687,
      "epoch": 0.8473154362416108,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004935206781596799,
      "loss": 5.5189,
      "mean_token_accuracy": 0.19375313222408294,
      "num_tokens": 17875314.0,
      "step": 9595
    },
    {
      "entropy": 5.884050559997559,
      "epoch": 0.847756976333451,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004935131730594315,
      "loss": 5.5732,
      "mean_token_accuracy": 0.1926530808210373,
      "num_tokens": 17884474.0,
      "step": 9600
    },
    {
      "entropy": 5.906814432144165,
      "epoch": 0.8481985164252914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004935056636786342,
      "loss": 5.6803,
      "mean_token_accuracy": 0.18539205491542815,
      "num_tokens": 17893386.0,
      "step": 9605
    },
    {
      "entropy": 5.972839593887329,
      "epoch": 0.8486400565171317,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004934981500174353,
      "loss": 5.6352,
      "mean_token_accuracy": 0.18344967812299728,
      "num_tokens": 17902564.0,
      "step": 9610
    },
    {
      "entropy": 5.907458353042602,
      "epoch": 0.8490815966089721,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934906320759818,
      "loss": 5.6706,
      "mean_token_accuracy": 0.17910434454679489,
      "num_tokens": 17911609.0,
      "step": 9615
    },
    {
      "entropy": 5.914002275466919,
      "epoch": 0.8495231367008125,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934831098544211,
      "loss": 5.6993,
      "mean_token_accuracy": 0.1844701811671257,
      "num_tokens": 17921469.0,
      "step": 9620
    },
    {
      "entropy": 5.90819149017334,
      "epoch": 0.8499646767926528,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004934755833529007,
      "loss": 5.5414,
      "mean_token_accuracy": 0.19303625375032424,
      "num_tokens": 17930709.0,
      "step": 9625
    },
    {
      "entropy": 5.976983118057251,
      "epoch": 0.8504062168844931,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004934680525715677,
      "loss": 5.6313,
      "mean_token_accuracy": 0.18149446696043015,
      "num_tokens": 17939864.0,
      "step": 9630
    },
    {
      "entropy": 5.838002252578735,
      "epoch": 0.8508477569763334,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004934605175105698,
      "loss": 5.5037,
      "mean_token_accuracy": 0.196158330142498,
      "num_tokens": 17949302.0,
      "step": 9635
    },
    {
      "entropy": 5.87263765335083,
      "epoch": 0.8512892970681738,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004934529781700546,
      "loss": 5.6156,
      "mean_token_accuracy": 0.18246717751026154,
      "num_tokens": 17959191.0,
      "step": 9640
    },
    {
      "entropy": 5.9224142074584964,
      "epoch": 0.8517308371600141,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934454345501698,
      "loss": 5.6201,
      "mean_token_accuracy": 0.18667593747377395,
      "num_tokens": 17967209.0,
      "step": 9645
    },
    {
      "entropy": 5.930680847167968,
      "epoch": 0.8521723772518545,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004934378866510633,
      "loss": 5.6165,
      "mean_token_accuracy": 0.18536110818386078,
      "num_tokens": 17976858.0,
      "step": 9650
    },
    {
      "entropy": 5.939687061309814,
      "epoch": 0.8526139173436948,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934303344728828,
      "loss": 5.5814,
      "mean_token_accuracy": 0.18879270553588867,
      "num_tokens": 17986059.0,
      "step": 9655
    },
    {
      "entropy": 5.93117504119873,
      "epoch": 0.8530554574355351,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004934227780157761,
      "loss": 5.6482,
      "mean_token_accuracy": 0.1911301165819168,
      "num_tokens": 17994765.0,
      "step": 9660
    },
    {
      "entropy": 5.930541133880615,
      "epoch": 0.8534969975273755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004934152172798916,
      "loss": 5.5797,
      "mean_token_accuracy": 0.1837850660085678,
      "num_tokens": 18003801.0,
      "step": 9665
    },
    {
      "entropy": 5.968143320083618,
      "epoch": 0.8539385376192158,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004934076522653772,
      "loss": 5.6537,
      "mean_token_accuracy": 0.179361928999424,
      "num_tokens": 18012905.0,
      "step": 9670
    },
    {
      "entropy": 5.8088632106781,
      "epoch": 0.8543800777110562,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000493400082972381,
      "loss": 5.5618,
      "mean_token_accuracy": 0.19472831934690477,
      "num_tokens": 18023137.0,
      "step": 9675
    },
    {
      "entropy": 5.916323471069336,
      "epoch": 0.8548216178028964,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933925094010516,
      "loss": 5.7464,
      "mean_token_accuracy": 0.18209900110960006,
      "num_tokens": 18032469.0,
      "step": 9680
    },
    {
      "entropy": 6.012365913391113,
      "epoch": 0.8552631578947368,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004933849315515369,
      "loss": 5.6905,
      "mean_token_accuracy": 0.17912546247243882,
      "num_tokens": 18042486.0,
      "step": 9685
    },
    {
      "entropy": 6.038495063781738,
      "epoch": 0.8557046979865772,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004933773494239857,
      "loss": 5.6741,
      "mean_token_accuracy": 0.18551600426435472,
      "num_tokens": 18051239.0,
      "step": 9690
    },
    {
      "entropy": 6.016125917434692,
      "epoch": 0.8561462380784175,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004933697630185464,
      "loss": 5.7503,
      "mean_token_accuracy": 0.18361809849739075,
      "num_tokens": 18061082.0,
      "step": 9695
    },
    {
      "entropy": 5.895164728164673,
      "epoch": 0.8565877781702579,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933621723353677,
      "loss": 5.6455,
      "mean_token_accuracy": 0.18684351295232773,
      "num_tokens": 18070713.0,
      "step": 9700
    },
    {
      "entropy": 5.974310827255249,
      "epoch": 0.8570293182620982,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493354577374598,
      "loss": 5.6445,
      "mean_token_accuracy": 0.1842496320605278,
      "num_tokens": 18080182.0,
      "step": 9705
    },
    {
      "entropy": 5.961086797714233,
      "epoch": 0.8574708583539385,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004933469781363865,
      "loss": 5.6399,
      "mean_token_accuracy": 0.18605689108371734,
      "num_tokens": 18089638.0,
      "step": 9710
    },
    {
      "entropy": 5.885256767272949,
      "epoch": 0.8579123984457788,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004933393746208819,
      "loss": 5.5285,
      "mean_token_accuracy": 0.19568028151988984,
      "num_tokens": 18098473.0,
      "step": 9715
    },
    {
      "entropy": 5.867877721786499,
      "epoch": 0.8583539385376192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933317668282331,
      "loss": 5.6592,
      "mean_token_accuracy": 0.1828484743833542,
      "num_tokens": 18107378.0,
      "step": 9720
    },
    {
      "entropy": 5.912347745895386,
      "epoch": 0.8587954786294596,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004933241547585891,
      "loss": 5.5599,
      "mean_token_accuracy": 0.1929926872253418,
      "num_tokens": 18117191.0,
      "step": 9725
    },
    {
      "entropy": 6.033688688278199,
      "epoch": 0.8592370187212999,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004933165384120991,
      "loss": 5.7993,
      "mean_token_accuracy": 0.17835478335618973,
      "num_tokens": 18128169.0,
      "step": 9730
    },
    {
      "entropy": 5.858069133758545,
      "epoch": 0.8596785588131403,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933089177889124,
      "loss": 5.4641,
      "mean_token_accuracy": 0.19777891635894776,
      "num_tokens": 18136703.0,
      "step": 9735
    },
    {
      "entropy": 5.877928876876831,
      "epoch": 0.8601200989049805,
      "grad_norm": 0.875,
      "learning_rate": 0.0004933012928891781,
      "loss": 5.6357,
      "mean_token_accuracy": 0.1937806323170662,
      "num_tokens": 18145691.0,
      "step": 9740
    },
    {
      "entropy": 5.881228828430176,
      "epoch": 0.8605616389968209,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932936637130456,
      "loss": 5.5905,
      "mean_token_accuracy": 0.1909533053636551,
      "num_tokens": 18154744.0,
      "step": 9745
    },
    {
      "entropy": 5.966592836380005,
      "epoch": 0.8610031790886612,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004932860302606644,
      "loss": 5.6404,
      "mean_token_accuracy": 0.19115128815174104,
      "num_tokens": 18164506.0,
      "step": 9750
    },
    {
      "entropy": 5.883775854110718,
      "epoch": 0.8614447191805016,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004932783925321838,
      "loss": 5.4479,
      "mean_token_accuracy": 0.20141887068748474,
      "num_tokens": 18172841.0,
      "step": 9755
    },
    {
      "entropy": 5.8807713985443115,
      "epoch": 0.861886259272342,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493270750527754,
      "loss": 5.5951,
      "mean_token_accuracy": 0.18641975224018098,
      "num_tokens": 18182055.0,
      "step": 9760
    },
    {
      "entropy": 5.969906616210937,
      "epoch": 0.8623277993641822,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932631042475241,
      "loss": 5.6195,
      "mean_token_accuracy": 0.18463124334812164,
      "num_tokens": 18191213.0,
      "step": 9765
    },
    {
      "entropy": 5.834154510498047,
      "epoch": 0.8627693394560226,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932554536916441,
      "loss": 5.6053,
      "mean_token_accuracy": 0.18940655887126923,
      "num_tokens": 18200448.0,
      "step": 9770
    },
    {
      "entropy": 5.975690793991089,
      "epoch": 0.8632108795478629,
      "grad_norm": 0.875,
      "learning_rate": 0.000493247798860264,
      "loss": 5.6074,
      "mean_token_accuracy": 0.18649885505437852,
      "num_tokens": 18210262.0,
      "step": 9775
    },
    {
      "entropy": 5.949877834320068,
      "epoch": 0.8636524196397033,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932401397535337,
      "loss": 5.5969,
      "mean_token_accuracy": 0.1910925716161728,
      "num_tokens": 18219314.0,
      "step": 9780
    },
    {
      "entropy": 5.9349922180175785,
      "epoch": 0.8640939597315436,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493232476371603,
      "loss": 5.6643,
      "mean_token_accuracy": 0.1890694186091423,
      "num_tokens": 18229247.0,
      "step": 9785
    },
    {
      "entropy": 5.889355230331421,
      "epoch": 0.864535499823384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004932248087146223,
      "loss": 5.6699,
      "mean_token_accuracy": 0.17438826858997344,
      "num_tokens": 18238871.0,
      "step": 9790
    },
    {
      "entropy": 6.006150484085083,
      "epoch": 0.8649770399152243,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932171367827417,
      "loss": 5.6617,
      "mean_token_accuracy": 0.1899381920695305,
      "num_tokens": 18248007.0,
      "step": 9795
    },
    {
      "entropy": 5.919534969329834,
      "epoch": 0.8654185800070646,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004932094605761115,
      "loss": 5.5259,
      "mean_token_accuracy": 0.19411712288856506,
      "num_tokens": 18257296.0,
      "step": 9800
    },
    {
      "entropy": 5.89392409324646,
      "epoch": 0.865860120098905,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004932017800948821,
      "loss": 5.6422,
      "mean_token_accuracy": 0.18463580459356307,
      "num_tokens": 18267513.0,
      "step": 9805
    },
    {
      "entropy": 5.910326242446899,
      "epoch": 0.8663016601907453,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004931940953392038,
      "loss": 5.6413,
      "mean_token_accuracy": 0.1811020031571388,
      "num_tokens": 18276331.0,
      "step": 9810
    },
    {
      "entropy": 5.916569232940674,
      "epoch": 0.8667432002825857,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004931864063092274,
      "loss": 5.5904,
      "mean_token_accuracy": 0.18913673013448715,
      "num_tokens": 18285368.0,
      "step": 9815
    },
    {
      "entropy": 5.908625507354737,
      "epoch": 0.8671847403744259,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004931787130051034,
      "loss": 5.6139,
      "mean_token_accuracy": 0.18909970670938492,
      "num_tokens": 18294254.0,
      "step": 9820
    },
    {
      "entropy": 5.953518962860107,
      "epoch": 0.8676262804662663,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931710154269824,
      "loss": 5.5497,
      "mean_token_accuracy": 0.18732893615961074,
      "num_tokens": 18303275.0,
      "step": 9825
    },
    {
      "entropy": 5.9172539710998535,
      "epoch": 0.8680678205581067,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004931633135750154,
      "loss": 5.6349,
      "mean_token_accuracy": 0.18304867893457413,
      "num_tokens": 18312284.0,
      "step": 9830
    },
    {
      "entropy": 5.952906703948974,
      "epoch": 0.868509360649947,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004931556074493531,
      "loss": 5.6656,
      "mean_token_accuracy": 0.18271292001008987,
      "num_tokens": 18321621.0,
      "step": 9835
    },
    {
      "entropy": 5.797186279296875,
      "epoch": 0.8689509007417874,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931478970501465,
      "loss": 5.4624,
      "mean_token_accuracy": 0.19594074338674544,
      "num_tokens": 18330472.0,
      "step": 9840
    },
    {
      "entropy": 5.91771936416626,
      "epoch": 0.8693924408336277,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931401823775466,
      "loss": 5.5977,
      "mean_token_accuracy": 0.18701441287994386,
      "num_tokens": 18339558.0,
      "step": 9845
    },
    {
      "entropy": 5.918262767791748,
      "epoch": 0.869833980925468,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004931324634317047,
      "loss": 5.5761,
      "mean_token_accuracy": 0.19163611978292466,
      "num_tokens": 18348540.0,
      "step": 9850
    },
    {
      "entropy": 5.955055475234985,
      "epoch": 0.8702755210173083,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004931247402127719,
      "loss": 5.6906,
      "mean_token_accuracy": 0.176729616522789,
      "num_tokens": 18358436.0,
      "step": 9855
    },
    {
      "entropy": 5.936215543746949,
      "epoch": 0.8707170611091487,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004931170127208995,
      "loss": 5.5631,
      "mean_token_accuracy": 0.1893255516886711,
      "num_tokens": 18367030.0,
      "step": 9860
    },
    {
      "entropy": 5.917493724822998,
      "epoch": 0.8711586012009891,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004931092809562388,
      "loss": 5.6032,
      "mean_token_accuracy": 0.1922762483358383,
      "num_tokens": 18376514.0,
      "step": 9865
    },
    {
      "entropy": 5.982113313674927,
      "epoch": 0.8716001412928294,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931015449189414,
      "loss": 5.6722,
      "mean_token_accuracy": 0.18113065212965013,
      "num_tokens": 18385876.0,
      "step": 9870
    },
    {
      "entropy": 5.95506739616394,
      "epoch": 0.8720416813846698,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004930938046091587,
      "loss": 5.5981,
      "mean_token_accuracy": 0.1855507880449295,
      "num_tokens": 18394448.0,
      "step": 9875
    },
    {
      "entropy": 5.950230884552002,
      "epoch": 0.87248322147651,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930860600270425,
      "loss": 5.6711,
      "mean_token_accuracy": 0.18222594112157822,
      "num_tokens": 18404043.0,
      "step": 9880
    },
    {
      "entropy": 5.903411436080932,
      "epoch": 0.8729247615683504,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930783111727443,
      "loss": 5.5225,
      "mean_token_accuracy": 0.18824177533388137,
      "num_tokens": 18412458.0,
      "step": 9885
    },
    {
      "entropy": 5.855086708068848,
      "epoch": 0.8733663016601907,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493070558046416,
      "loss": 5.6537,
      "mean_token_accuracy": 0.18128291368484498,
      "num_tokens": 18421489.0,
      "step": 9890
    },
    {
      "entropy": 5.953315353393554,
      "epoch": 0.8738078417520311,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004930628006482097,
      "loss": 5.6147,
      "mean_token_accuracy": 0.18333493024110795,
      "num_tokens": 18430197.0,
      "step": 9895
    },
    {
      "entropy": 6.009244203567505,
      "epoch": 0.8742493818438715,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004930550389782769,
      "loss": 5.6414,
      "mean_token_accuracy": 0.18415736109018327,
      "num_tokens": 18439498.0,
      "step": 9900
    },
    {
      "entropy": 5.88096513748169,
      "epoch": 0.8746909219357117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930472730367701,
      "loss": 5.6713,
      "mean_token_accuracy": 0.1895434468984604,
      "num_tokens": 18449528.0,
      "step": 9905
    },
    {
      "entropy": 5.9174731254577635,
      "epoch": 0.8751324620275521,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004930395028238412,
      "loss": 5.5815,
      "mean_token_accuracy": 0.1863342747092247,
      "num_tokens": 18458687.0,
      "step": 9910
    },
    {
      "entropy": 5.950844430923462,
      "epoch": 0.8755740021193924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004930317283396423,
      "loss": 5.6624,
      "mean_token_accuracy": 0.19057082682847976,
      "num_tokens": 18468368.0,
      "step": 9915
    },
    {
      "entropy": 5.855986833572388,
      "epoch": 0.8760155422112328,
      "grad_norm": 1.0,
      "learning_rate": 0.000493023949584326,
      "loss": 5.6085,
      "mean_token_accuracy": 0.18856714516878129,
      "num_tokens": 18477973.0,
      "step": 9920
    },
    {
      "entropy": 5.913123655319214,
      "epoch": 0.8764570823030731,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004930161665580445,
      "loss": 5.4818,
      "mean_token_accuracy": 0.1948400303721428,
      "num_tokens": 18486823.0,
      "step": 9925
    },
    {
      "entropy": 5.930782318115234,
      "epoch": 0.8768986223949135,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004930083792609502,
      "loss": 5.5661,
      "mean_token_accuracy": 0.18187194615602492,
      "num_tokens": 18496032.0,
      "step": 9930
    },
    {
      "entropy": 5.799260568618775,
      "epoch": 0.8773401624867538,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004930005876931958,
      "loss": 5.541,
      "mean_token_accuracy": 0.19032247960567475,
      "num_tokens": 18504038.0,
      "step": 9935
    },
    {
      "entropy": 5.994828939437866,
      "epoch": 0.8777817025785941,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004929927918549339,
      "loss": 5.6799,
      "mean_token_accuracy": 0.17964600622653962,
      "num_tokens": 18513682.0,
      "step": 9940
    },
    {
      "entropy": 6.008097314834595,
      "epoch": 0.8782232426704345,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929849917463171,
      "loss": 5.6627,
      "mean_token_accuracy": 0.1785341516137123,
      "num_tokens": 18523744.0,
      "step": 9945
    },
    {
      "entropy": 5.927927541732788,
      "epoch": 0.8786647827622748,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004929771873674984,
      "loss": 5.5697,
      "mean_token_accuracy": 0.18869753032922745,
      "num_tokens": 18532649.0,
      "step": 9950
    },
    {
      "entropy": 5.8934080600738525,
      "epoch": 0.8791063228541152,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929693787186305,
      "loss": 5.5494,
      "mean_token_accuracy": 0.18713405430316926,
      "num_tokens": 18541866.0,
      "step": 9955
    },
    {
      "entropy": 5.948517608642578,
      "epoch": 0.8795478629459554,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004929615657998664,
      "loss": 5.5938,
      "mean_token_accuracy": 0.18316649049520492,
      "num_tokens": 18550597.0,
      "step": 9960
    },
    {
      "entropy": 5.855611371994018,
      "epoch": 0.8799894030377958,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929537486113593,
      "loss": 5.617,
      "mean_token_accuracy": 0.18819984495639802,
      "num_tokens": 18560059.0,
      "step": 9965
    },
    {
      "entropy": 5.9186090469360355,
      "epoch": 0.8804309431296362,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929459271532621,
      "loss": 5.597,
      "mean_token_accuracy": 0.18555559366941451,
      "num_tokens": 18568815.0,
      "step": 9970
    },
    {
      "entropy": 5.915519142150879,
      "epoch": 0.8808724832214765,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929381014257282,
      "loss": 5.6062,
      "mean_token_accuracy": 0.19008751660585405,
      "num_tokens": 18577919.0,
      "step": 9975
    },
    {
      "entropy": 5.948284149169922,
      "epoch": 0.8813140233133169,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004929302714289107,
      "loss": 5.6804,
      "mean_token_accuracy": 0.18061696887016296,
      "num_tokens": 18587800.0,
      "step": 9980
    },
    {
      "entropy": 5.987317419052124,
      "epoch": 0.8817555634051572,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004929224371629634,
      "loss": 5.6651,
      "mean_token_accuracy": 0.18760334700345993,
      "num_tokens": 18596954.0,
      "step": 9985
    },
    {
      "entropy": 5.936594820022583,
      "epoch": 0.8821971034969975,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004929145986280394,
      "loss": 5.6093,
      "mean_token_accuracy": 0.18292579799890518,
      "num_tokens": 18605945.0,
      "step": 9990
    },
    {
      "entropy": 5.936980056762695,
      "epoch": 0.8826386435888378,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929067558242922,
      "loss": 5.6615,
      "mean_token_accuracy": 0.1811128407716751,
      "num_tokens": 18615523.0,
      "step": 9995
    },
    {
      "entropy": 5.917059373855591,
      "epoch": 0.8830801836806782,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004928989087518758,
      "loss": 5.5964,
      "mean_token_accuracy": 0.18397147655487062,
      "num_tokens": 18625217.0,
      "step": 10000
    },
    {
      "entropy": 5.896638536453247,
      "epoch": 0.8835217237725186,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004928910574109435,
      "loss": 5.5881,
      "mean_token_accuracy": 0.19273042529821396,
      "num_tokens": 18633444.0,
      "step": 10005
    },
    {
      "entropy": 5.833864307403564,
      "epoch": 0.8839632638643589,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004928832018016495,
      "loss": 5.5786,
      "mean_token_accuracy": 0.19139365404844283,
      "num_tokens": 18643449.0,
      "step": 10010
    },
    {
      "entropy": 5.862657642364502,
      "epoch": 0.8844048039561992,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004928753419241472,
      "loss": 5.5427,
      "mean_token_accuracy": 0.19022172391414643,
      "num_tokens": 18651372.0,
      "step": 10015
    },
    {
      "entropy": 5.945731687545776,
      "epoch": 0.8848463440480395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928674777785912,
      "loss": 5.648,
      "mean_token_accuracy": 0.18019123375415802,
      "num_tokens": 18660777.0,
      "step": 10020
    },
    {
      "entropy": 5.975177764892578,
      "epoch": 0.8852878841398799,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004928596093651351,
      "loss": 5.8219,
      "mean_token_accuracy": 0.16846104860305786,
      "num_tokens": 18671384.0,
      "step": 10025
    },
    {
      "entropy": 5.9661290645599365,
      "epoch": 0.8857294242317202,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492851736683933,
      "loss": 5.6331,
      "mean_token_accuracy": 0.1818322941660881,
      "num_tokens": 18680399.0,
      "step": 10030
    },
    {
      "entropy": 5.988116121292114,
      "epoch": 0.8861709643235606,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004928438597351395,
      "loss": 5.6208,
      "mean_token_accuracy": 0.18413838297128676,
      "num_tokens": 18690986.0,
      "step": 10035
    },
    {
      "entropy": 5.9103940486907955,
      "epoch": 0.886612504415401,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004928359785189086,
      "loss": 5.6211,
      "mean_token_accuracy": 0.1892377406358719,
      "num_tokens": 18700006.0,
      "step": 10040
    },
    {
      "entropy": 5.863557481765747,
      "epoch": 0.8870540445072412,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928280930353948,
      "loss": 5.5011,
      "mean_token_accuracy": 0.19607726633548736,
      "num_tokens": 18708473.0,
      "step": 10045
    },
    {
      "entropy": 5.876119184494018,
      "epoch": 0.8874955845990816,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004928202032847525,
      "loss": 5.5525,
      "mean_token_accuracy": 0.18564749211072923,
      "num_tokens": 18717309.0,
      "step": 10050
    },
    {
      "entropy": 5.867317771911621,
      "epoch": 0.8879371246909219,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928123092671362,
      "loss": 5.5583,
      "mean_token_accuracy": 0.18877305537462236,
      "num_tokens": 18726611.0,
      "step": 10055
    },
    {
      "entropy": 5.9733811378479,
      "epoch": 0.8883786647827623,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004928044109827008,
      "loss": 5.6741,
      "mean_token_accuracy": 0.18315087556838988,
      "num_tokens": 18737054.0,
      "step": 10060
    },
    {
      "entropy": 5.933556699752808,
      "epoch": 0.8888202048746026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004927965084316007,
      "loss": 5.6015,
      "mean_token_accuracy": 0.1835976481437683,
      "num_tokens": 18745595.0,
      "step": 10065
    },
    {
      "entropy": 5.935288286209106,
      "epoch": 0.889261744966443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000492788601613991,
      "loss": 5.5693,
      "mean_token_accuracy": 0.18996552377939224,
      "num_tokens": 18754306.0,
      "step": 10070
    },
    {
      "entropy": 5.928145027160644,
      "epoch": 0.8897032850582833,
      "grad_norm": 3.375,
      "learning_rate": 0.0004927806905300265,
      "loss": 5.7495,
      "mean_token_accuracy": 0.18096244260668753,
      "num_tokens": 18764344.0,
      "step": 10075
    },
    {
      "entropy": 5.965080976486206,
      "epoch": 0.8901448251501236,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492772775179862,
      "loss": 5.6375,
      "mean_token_accuracy": 0.18465467542409897,
      "num_tokens": 18773775.0,
      "step": 10080
    },
    {
      "entropy": 5.870256090164185,
      "epoch": 0.890586365241964,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004927648555636528,
      "loss": 5.5069,
      "mean_token_accuracy": 0.1890522673726082,
      "num_tokens": 18782432.0,
      "step": 10085
    },
    {
      "entropy": 5.881683397293091,
      "epoch": 0.8910279053338043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004927569316815539,
      "loss": 5.6493,
      "mean_token_accuracy": 0.18701877743005751,
      "num_tokens": 18792070.0,
      "step": 10090
    },
    {
      "entropy": 5.933225059509278,
      "epoch": 0.8914694454256447,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927490035337205,
      "loss": 5.6008,
      "mean_token_accuracy": 0.1833445981144905,
      "num_tokens": 18801282.0,
      "step": 10095
    },
    {
      "entropy": 5.960941505432129,
      "epoch": 0.8919109855174849,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000492741071120308,
      "loss": 5.6757,
      "mean_token_accuracy": 0.1833571344614029,
      "num_tokens": 18811232.0,
      "step": 10100
    },
    {
      "entropy": 5.940049839019776,
      "epoch": 0.8923525256093253,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927331344414717,
      "loss": 5.6177,
      "mean_token_accuracy": 0.17775194644927977,
      "num_tokens": 18820180.0,
      "step": 10105
    },
    {
      "entropy": 5.882258605957031,
      "epoch": 0.8927940657011657,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004927251934973672,
      "loss": 5.5623,
      "mean_token_accuracy": 0.19483665078878404,
      "num_tokens": 18829173.0,
      "step": 10110
    },
    {
      "entropy": 5.951724529266357,
      "epoch": 0.893235605793006,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004927172482881499,
      "loss": 5.5593,
      "mean_token_accuracy": 0.1943387970328331,
      "num_tokens": 18838704.0,
      "step": 10115
    },
    {
      "entropy": 5.9025708675384525,
      "epoch": 0.8936771458848464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004927092988139755,
      "loss": 5.5432,
      "mean_token_accuracy": 0.20021082311868668,
      "num_tokens": 18847246.0,
      "step": 10120
    },
    {
      "entropy": 5.888104343414307,
      "epoch": 0.8941186859766866,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004927013450749999,
      "loss": 5.7196,
      "mean_token_accuracy": 0.1836605727672577,
      "num_tokens": 18856612.0,
      "step": 10125
    },
    {
      "entropy": 5.926151609420776,
      "epoch": 0.894560226068527,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926933870713787,
      "loss": 5.5812,
      "mean_token_accuracy": 0.1938813269138336,
      "num_tokens": 18866419.0,
      "step": 10130
    },
    {
      "entropy": 6.013955688476562,
      "epoch": 0.8950017661603673,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926854248032678,
      "loss": 5.6905,
      "mean_token_accuracy": 0.18183711767196656,
      "num_tokens": 18875747.0,
      "step": 10135
    },
    {
      "entropy": 5.900475358963012,
      "epoch": 0.8954433062522077,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004926774582708234,
      "loss": 5.631,
      "mean_token_accuracy": 0.18517805337905885,
      "num_tokens": 18884417.0,
      "step": 10140
    },
    {
      "entropy": 5.872224378585815,
      "epoch": 0.8958848463440481,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004926694874742012,
      "loss": 5.5944,
      "mean_token_accuracy": 0.18872942477464677,
      "num_tokens": 18893432.0,
      "step": 10145
    },
    {
      "entropy": 5.937736511230469,
      "epoch": 0.8963263864358884,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926615124135577,
      "loss": 5.6375,
      "mean_token_accuracy": 0.1809471681714058,
      "num_tokens": 18902903.0,
      "step": 10150
    },
    {
      "entropy": 5.865172815322876,
      "epoch": 0.8967679265277287,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926535330890488,
      "loss": 5.488,
      "mean_token_accuracy": 0.19685240387916564,
      "num_tokens": 18912318.0,
      "step": 10155
    },
    {
      "entropy": 5.992293214797973,
      "epoch": 0.897209466619569,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004926455495008311,
      "loss": 5.7144,
      "mean_token_accuracy": 0.17530783712863923,
      "num_tokens": 18922768.0,
      "step": 10160
    },
    {
      "entropy": 5.915782737731933,
      "epoch": 0.8976510067114094,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004926375616490608,
      "loss": 5.6311,
      "mean_token_accuracy": 0.18454720377922057,
      "num_tokens": 18932694.0,
      "step": 10165
    },
    {
      "entropy": 5.895670938491821,
      "epoch": 0.8980925468032497,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926295695338945,
      "loss": 5.6505,
      "mean_token_accuracy": 0.18947786390781401,
      "num_tokens": 18942714.0,
      "step": 10170
    },
    {
      "entropy": 6.003636693954467,
      "epoch": 0.8985340868950901,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926215731554887,
      "loss": 5.6474,
      "mean_token_accuracy": 0.1847643792629242,
      "num_tokens": 18953241.0,
      "step": 10175
    },
    {
      "entropy": 5.967898607254028,
      "epoch": 0.8989756269869305,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492613572514,
      "loss": 5.5592,
      "mean_token_accuracy": 0.19308606088161467,
      "num_tokens": 18962822.0,
      "step": 10180
    },
    {
      "entropy": 5.96004285812378,
      "epoch": 0.8994171670787707,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926055676095851,
      "loss": 5.6629,
      "mean_token_accuracy": 0.1846911773085594,
      "num_tokens": 18972267.0,
      "step": 10185
    },
    {
      "entropy": 5.909711837768555,
      "epoch": 0.8998587071706111,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004925975584424012,
      "loss": 5.6226,
      "mean_token_accuracy": 0.1863635554909706,
      "num_tokens": 18981342.0,
      "step": 10190
    },
    {
      "entropy": 5.908374166488647,
      "epoch": 0.9003002472624514,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004925895450126046,
      "loss": 5.553,
      "mean_token_accuracy": 0.19136232137680054,
      "num_tokens": 18989818.0,
      "step": 10195
    },
    {
      "entropy": 5.9588623046875,
      "epoch": 0.9007417873542918,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925815273203526,
      "loss": 5.734,
      "mean_token_accuracy": 0.1758061945438385,
      "num_tokens": 18998542.0,
      "step": 10200
    },
    {
      "entropy": 5.9603354930877686,
      "epoch": 0.901183327446132,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004925735053658022,
      "loss": 5.642,
      "mean_token_accuracy": 0.18448109477758406,
      "num_tokens": 19008050.0,
      "step": 10205
    },
    {
      "entropy": 5.914842224121093,
      "epoch": 0.9016248675379724,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004925654791491106,
      "loss": 5.6029,
      "mean_token_accuracy": 0.18933584541082382,
      "num_tokens": 19017762.0,
      "step": 10210
    },
    {
      "entropy": 5.950772619247436,
      "epoch": 0.9020664076298128,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925574486704351,
      "loss": 5.6344,
      "mean_token_accuracy": 0.18237055391073226,
      "num_tokens": 19027172.0,
      "step": 10215
    },
    {
      "entropy": 5.951493167877198,
      "epoch": 0.9025079477216531,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004925494139299329,
      "loss": 5.7332,
      "mean_token_accuracy": 0.17888642847537994,
      "num_tokens": 19036858.0,
      "step": 10220
    },
    {
      "entropy": 5.8613263130187985,
      "epoch": 0.9029494878134935,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925413749277613,
      "loss": 5.5459,
      "mean_token_accuracy": 0.18829281032085418,
      "num_tokens": 19045653.0,
      "step": 10225
    },
    {
      "entropy": 5.825928735733032,
      "epoch": 0.9033910279053338,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004925333316640779,
      "loss": 5.5232,
      "mean_token_accuracy": 0.1929013654589653,
      "num_tokens": 19055945.0,
      "step": 10230
    },
    {
      "entropy": 5.9103076457977295,
      "epoch": 0.9038325679971742,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004925252841390403,
      "loss": 5.7267,
      "mean_token_accuracy": 0.18506952673196791,
      "num_tokens": 19065239.0,
      "step": 10235
    },
    {
      "entropy": 6.022396278381348,
      "epoch": 0.9042741080890144,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004925172323528061,
      "loss": 5.6271,
      "mean_token_accuracy": 0.18314308077096939,
      "num_tokens": 19074384.0,
      "step": 10240
    },
    {
      "entropy": 5.934940671920776,
      "epoch": 0.9047156481808548,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492509176305533,
      "loss": 5.7055,
      "mean_token_accuracy": 0.17762739062309266,
      "num_tokens": 19084022.0,
      "step": 10245
    },
    {
      "entropy": 5.987552881240845,
      "epoch": 0.9051571882726952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925011159973788,
      "loss": 5.669,
      "mean_token_accuracy": 0.18072815090417862,
      "num_tokens": 19093873.0,
      "step": 10250
    },
    {
      "entropy": 5.904093265533447,
      "epoch": 0.9055987283645355,
      "grad_norm": 0.875,
      "learning_rate": 0.0004924930514285015,
      "loss": 5.6152,
      "mean_token_accuracy": 0.1832732766866684,
      "num_tokens": 19103266.0,
      "step": 10255
    },
    {
      "entropy": 5.934785461425781,
      "epoch": 0.9060402684563759,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000492484982599059,
      "loss": 5.5661,
      "mean_token_accuracy": 0.18928578794002532,
      "num_tokens": 19112705.0,
      "step": 10260
    },
    {
      "entropy": 5.837735366821289,
      "epoch": 0.9064818085482161,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004924769095092094,
      "loss": 5.5441,
      "mean_token_accuracy": 0.19110118895769118,
      "num_tokens": 19121321.0,
      "step": 10265
    },
    {
      "entropy": 5.904051971435547,
      "epoch": 0.9069233486400565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004924688321591109,
      "loss": 5.5854,
      "mean_token_accuracy": 0.18986263126134872,
      "num_tokens": 19131637.0,
      "step": 10270
    },
    {
      "entropy": 5.962698030471802,
      "epoch": 0.9073648887318968,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004924607505489215,
      "loss": 5.5769,
      "mean_token_accuracy": 0.1841653034090996,
      "num_tokens": 19141164.0,
      "step": 10275
    },
    {
      "entropy": 5.910167026519775,
      "epoch": 0.9078064288237372,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004924526646787998,
      "loss": 5.6903,
      "mean_token_accuracy": 0.18670402467250824,
      "num_tokens": 19151369.0,
      "step": 10280
    },
    {
      "entropy": 5.879006004333496,
      "epoch": 0.9082479689155776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000492444574548904,
      "loss": 5.5552,
      "mean_token_accuracy": 0.19237009733915328,
      "num_tokens": 19160710.0,
      "step": 10285
    },
    {
      "entropy": 5.8453813076019285,
      "epoch": 0.9086895090074179,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924364801593927,
      "loss": 5.4836,
      "mean_token_accuracy": 0.19855196326971053,
      "num_tokens": 19169801.0,
      "step": 10290
    },
    {
      "entropy": 5.810707330703735,
      "epoch": 0.9091310490992582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924283815104243,
      "loss": 5.4783,
      "mean_token_accuracy": 0.19353894144296646,
      "num_tokens": 19178429.0,
      "step": 10295
    },
    {
      "entropy": 5.884520626068115,
      "epoch": 0.9095725891910985,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004924202786021578,
      "loss": 5.6197,
      "mean_token_accuracy": 0.18204735666513444,
      "num_tokens": 19188061.0,
      "step": 10300
    },
    {
      "entropy": 6.008539152145386,
      "epoch": 0.9100141292829389,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004924121714347515,
      "loss": 5.6477,
      "mean_token_accuracy": 0.1814591482281685,
      "num_tokens": 19197787.0,
      "step": 10305
    },
    {
      "entropy": 5.959963607788086,
      "epoch": 0.9104556693747792,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924040600083644,
      "loss": 5.7195,
      "mean_token_accuracy": 0.1779332011938095,
      "num_tokens": 19208090.0,
      "step": 10310
    },
    {
      "entropy": 5.977293586730957,
      "epoch": 0.9108972094666196,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923959443231554,
      "loss": 5.7076,
      "mean_token_accuracy": 0.18379490375518798,
      "num_tokens": 19218067.0,
      "step": 10315
    },
    {
      "entropy": 5.948866701126098,
      "epoch": 0.91133874955846,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004923878243792836,
      "loss": 5.5981,
      "mean_token_accuracy": 0.18710697889328004,
      "num_tokens": 19227753.0,
      "step": 10320
    },
    {
      "entropy": 5.880122947692871,
      "epoch": 0.9117802896503002,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004923797001769079,
      "loss": 5.5766,
      "mean_token_accuracy": 0.1902680665254593,
      "num_tokens": 19237425.0,
      "step": 10325
    },
    {
      "entropy": 5.88461446762085,
      "epoch": 0.9122218297421406,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923715717161875,
      "loss": 5.6037,
      "mean_token_accuracy": 0.18889866769313812,
      "num_tokens": 19247243.0,
      "step": 10330
    },
    {
      "entropy": 5.945171737670899,
      "epoch": 0.9126633698339809,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004923634389972817,
      "loss": 5.637,
      "mean_token_accuracy": 0.1847313240170479,
      "num_tokens": 19256837.0,
      "step": 10335
    },
    {
      "entropy": 5.964085292816162,
      "epoch": 0.9131049099258213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923553020203497,
      "loss": 5.6909,
      "mean_token_accuracy": 0.18374738693237305,
      "num_tokens": 19266286.0,
      "step": 10340
    },
    {
      "entropy": 5.977098894119263,
      "epoch": 0.9135464500176615,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923471607855511,
      "loss": 5.6326,
      "mean_token_accuracy": 0.182721845805645,
      "num_tokens": 19275324.0,
      "step": 10345
    },
    {
      "entropy": 5.9194732189178465,
      "epoch": 0.9139879901095019,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492339015293045,
      "loss": 5.7015,
      "mean_token_accuracy": 0.18361707478761674,
      "num_tokens": 19284650.0,
      "step": 10350
    },
    {
      "entropy": 5.925259971618653,
      "epoch": 0.9144295302013423,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923308655429913,
      "loss": 5.6441,
      "mean_token_accuracy": 0.18201907128095626,
      "num_tokens": 19293648.0,
      "step": 10355
    },
    {
      "entropy": 5.975694990158081,
      "epoch": 0.9148710702931826,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004923227115355495,
      "loss": 5.6495,
      "mean_token_accuracy": 0.18005977123975753,
      "num_tokens": 19302802.0,
      "step": 10360
    },
    {
      "entropy": 5.874110460281372,
      "epoch": 0.915312610385023,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004923145532708794,
      "loss": 5.6229,
      "mean_token_accuracy": 0.17879916876554489,
      "num_tokens": 19312028.0,
      "step": 10365
    },
    {
      "entropy": 5.949617099761963,
      "epoch": 0.9157541504768633,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004923063907491408,
      "loss": 5.5923,
      "mean_token_accuracy": 0.1848357155919075,
      "num_tokens": 19321308.0,
      "step": 10370
    },
    {
      "entropy": 5.947924375534058,
      "epoch": 0.9161956905687036,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004922982239704936,
      "loss": 5.6814,
      "mean_token_accuracy": 0.18527669161558152,
      "num_tokens": 19331982.0,
      "step": 10375
    },
    {
      "entropy": 5.889736461639404,
      "epoch": 0.9166372306605439,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004922900529350978,
      "loss": 5.5756,
      "mean_token_accuracy": 0.18363943547010422,
      "num_tokens": 19341994.0,
      "step": 10380
    },
    {
      "entropy": 5.952250957489014,
      "epoch": 0.9170787707523843,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922818776431134,
      "loss": 5.5547,
      "mean_token_accuracy": 0.18902692794799805,
      "num_tokens": 19350881.0,
      "step": 10385
    },
    {
      "entropy": 5.878201293945312,
      "epoch": 0.9175203108442247,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922736980947005,
      "loss": 5.5564,
      "mean_token_accuracy": 0.1979235291481018,
      "num_tokens": 19360722.0,
      "step": 10390
    },
    {
      "entropy": 5.885159301757812,
      "epoch": 0.917961850936065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004922655142900196,
      "loss": 5.597,
      "mean_token_accuracy": 0.18756526708602905,
      "num_tokens": 19369629.0,
      "step": 10395
    },
    {
      "entropy": 5.893330335617065,
      "epoch": 0.9184033910279054,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004922573262292308,
      "loss": 5.6566,
      "mean_token_accuracy": 0.18318665623664857,
      "num_tokens": 19378644.0,
      "step": 10400
    },
    {
      "entropy": 5.9086466312408445,
      "epoch": 0.9188449311197456,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004922491339124945,
      "loss": 5.4661,
      "mean_token_accuracy": 0.1916213646531105,
      "num_tokens": 19387414.0,
      "step": 10405
    },
    {
      "entropy": 5.939161920547486,
      "epoch": 0.919286471211586,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004922409373399713,
      "loss": 5.6752,
      "mean_token_accuracy": 0.18403153568506242,
      "num_tokens": 19396756.0,
      "step": 10410
    },
    {
      "entropy": 5.9333411693573,
      "epoch": 0.9197280113034263,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922327365118216,
      "loss": 5.5947,
      "mean_token_accuracy": 0.18974966555833817,
      "num_tokens": 19406074.0,
      "step": 10415
    },
    {
      "entropy": 5.934563541412354,
      "epoch": 0.9201695513952667,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922245314282062,
      "loss": 5.6381,
      "mean_token_accuracy": 0.18668539822101593,
      "num_tokens": 19416047.0,
      "step": 10420
    },
    {
      "entropy": 5.882915592193603,
      "epoch": 0.9206110914871071,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922163220892858,
      "loss": 5.5638,
      "mean_token_accuracy": 0.18855049908161164,
      "num_tokens": 19425542.0,
      "step": 10425
    },
    {
      "entropy": 5.953513050079346,
      "epoch": 0.9210526315789473,
      "grad_norm": 0.828125,
      "learning_rate": 0.000492208108495221,
      "loss": 5.7003,
      "mean_token_accuracy": 0.17710556238889694,
      "num_tokens": 19435154.0,
      "step": 10430
    },
    {
      "entropy": 5.884162187576294,
      "epoch": 0.9214941716707877,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492199890646173,
      "loss": 5.5485,
      "mean_token_accuracy": 0.1883462995290756,
      "num_tokens": 19443746.0,
      "step": 10435
    },
    {
      "entropy": 5.919364929199219,
      "epoch": 0.921935711762628,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004921916685423028,
      "loss": 5.6227,
      "mean_token_accuracy": 0.1876997247338295,
      "num_tokens": 19452225.0,
      "step": 10440
    },
    {
      "entropy": 5.954281663894653,
      "epoch": 0.9223772518544684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921834421837711,
      "loss": 5.6123,
      "mean_token_accuracy": 0.18744943290948868,
      "num_tokens": 19461475.0,
      "step": 10445
    },
    {
      "entropy": 5.914217138290406,
      "epoch": 0.9228187919463087,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921752115707394,
      "loss": 5.5285,
      "mean_token_accuracy": 0.1939025342464447,
      "num_tokens": 19469891.0,
      "step": 10450
    },
    {
      "entropy": 5.801268100738525,
      "epoch": 0.9232603320381491,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004921669767033688,
      "loss": 5.5449,
      "mean_token_accuracy": 0.1992613196372986,
      "num_tokens": 19478944.0,
      "step": 10455
    },
    {
      "entropy": 5.912322235107422,
      "epoch": 0.9237018721299894,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004921587375818207,
      "loss": 5.5989,
      "mean_token_accuracy": 0.18650982826948165,
      "num_tokens": 19487478.0,
      "step": 10460
    },
    {
      "entropy": 5.897990989685058,
      "epoch": 0.9241434122218297,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004921504942062564,
      "loss": 5.6028,
      "mean_token_accuracy": 0.19069903790950776,
      "num_tokens": 19496829.0,
      "step": 10465
    },
    {
      "entropy": 5.824184894561768,
      "epoch": 0.9245849523136701,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921422465768374,
      "loss": 5.5005,
      "mean_token_accuracy": 0.19380157738924025,
      "num_tokens": 19505762.0,
      "step": 10470
    },
    {
      "entropy": 5.914334487915039,
      "epoch": 0.9250264924055104,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921339946937253,
      "loss": 5.5989,
      "mean_token_accuracy": 0.19156884849071504,
      "num_tokens": 19514788.0,
      "step": 10475
    },
    {
      "entropy": 5.89178147315979,
      "epoch": 0.9254680324973508,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004921257385570818,
      "loss": 5.5211,
      "mean_token_accuracy": 0.19875165075063705,
      "num_tokens": 19524379.0,
      "step": 10480
    },
    {
      "entropy": 5.886852407455445,
      "epoch": 0.925909572589191,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921174781670685,
      "loss": 5.674,
      "mean_token_accuracy": 0.1850998044013977,
      "num_tokens": 19533452.0,
      "step": 10485
    },
    {
      "entropy": 5.967869853973388,
      "epoch": 0.9263511126810314,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004921092135238473,
      "loss": 5.6343,
      "mean_token_accuracy": 0.18229382783174514,
      "num_tokens": 19543318.0,
      "step": 10490
    },
    {
      "entropy": 5.874660968780518,
      "epoch": 0.9267926527728718,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049210094462758,
      "loss": 5.5509,
      "mean_token_accuracy": 0.18883130103349685,
      "num_tokens": 19552386.0,
      "step": 10495
    },
    {
      "entropy": 5.87246356010437,
      "epoch": 0.9272341928647121,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004920926714784288,
      "loss": 5.6124,
      "mean_token_accuracy": 0.18176539838314057,
      "num_tokens": 19561515.0,
      "step": 10500
    },
    {
      "entropy": 5.939922571182251,
      "epoch": 0.9276757329565525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004920843940765555,
      "loss": 5.6586,
      "mean_token_accuracy": 0.17937710732221604,
      "num_tokens": 19571674.0,
      "step": 10505
    },
    {
      "entropy": 5.966903877258301,
      "epoch": 0.9281172730483928,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004920761124221224,
      "loss": 5.5265,
      "mean_token_accuracy": 0.2036296769976616,
      "num_tokens": 19580532.0,
      "step": 10510
    },
    {
      "entropy": 5.9597978591918945,
      "epoch": 0.9285588131402331,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004920678265152919,
      "loss": 5.6792,
      "mean_token_accuracy": 0.1808280199766159,
      "num_tokens": 19589506.0,
      "step": 10515
    },
    {
      "entropy": 5.9044787883758545,
      "epoch": 0.9290003532320734,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000492059536356226,
      "loss": 5.6586,
      "mean_token_accuracy": 0.18035432547330857,
      "num_tokens": 19599705.0,
      "step": 10520
    },
    {
      "entropy": 5.989162492752075,
      "epoch": 0.9294418933239138,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920512419450872,
      "loss": 5.6517,
      "mean_token_accuracy": 0.18453942388296127,
      "num_tokens": 19609736.0,
      "step": 10525
    },
    {
      "entropy": 5.909270524978638,
      "epoch": 0.9298834334157542,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000492042943282038,
      "loss": 5.6101,
      "mean_token_accuracy": 0.19186623841524125,
      "num_tokens": 19619038.0,
      "step": 10530
    },
    {
      "entropy": 5.959709548950196,
      "epoch": 0.9303249735075945,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492034640367241,
      "loss": 5.702,
      "mean_token_accuracy": 0.18688322752714157,
      "num_tokens": 19628505.0,
      "step": 10535
    },
    {
      "entropy": 5.864881181716919,
      "epoch": 0.9307665135994349,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492026333200859,
      "loss": 5.5404,
      "mean_token_accuracy": 0.1992712914943695,
      "num_tokens": 19637157.0,
      "step": 10540
    },
    {
      "entropy": 5.870511054992676,
      "epoch": 0.9312080536912751,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920180217830543,
      "loss": 5.5778,
      "mean_token_accuracy": 0.18575091958045958,
      "num_tokens": 19646372.0,
      "step": 10545
    },
    {
      "entropy": 5.969358110427857,
      "epoch": 0.9316495937831155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049200970611399,
      "loss": 5.6707,
      "mean_token_accuracy": 0.18036315739154815,
      "num_tokens": 19655704.0,
      "step": 10550
    },
    {
      "entropy": 5.966890430450439,
      "epoch": 0.9320911338749558,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920013861938291,
      "loss": 5.5928,
      "mean_token_accuracy": 0.19134537130594254,
      "num_tokens": 19664802.0,
      "step": 10555
    },
    {
      "entropy": 5.927483367919922,
      "epoch": 0.9325326739667962,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919930620227345,
      "loss": 5.5718,
      "mean_token_accuracy": 0.19672607481479645,
      "num_tokens": 19674056.0,
      "step": 10560
    },
    {
      "entropy": 5.925628423690796,
      "epoch": 0.9329742140586366,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919847336008692,
      "loss": 5.692,
      "mean_token_accuracy": 0.18135405480861663,
      "num_tokens": 19684138.0,
      "step": 10565
    },
    {
      "entropy": 5.925083971023559,
      "epoch": 0.9334157541504768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919764009283963,
      "loss": 5.5607,
      "mean_token_accuracy": 0.18618064671754836,
      "num_tokens": 19692984.0,
      "step": 10570
    },
    {
      "entropy": 5.950464916229248,
      "epoch": 0.9338572942423172,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004919680640054792,
      "loss": 5.6532,
      "mean_token_accuracy": 0.183426134288311,
      "num_tokens": 19702321.0,
      "step": 10575
    },
    {
      "entropy": 5.91970043182373,
      "epoch": 0.9342988343341575,
      "grad_norm": 0.953125,
      "learning_rate": 0.000491959722832281,
      "loss": 5.556,
      "mean_token_accuracy": 0.19495438039302826,
      "num_tokens": 19710982.0,
      "step": 10580
    },
    {
      "entropy": 6.0049985408782955,
      "epoch": 0.9347403744259979,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004919513774089653,
      "loss": 5.7129,
      "mean_token_accuracy": 0.17703156918287277,
      "num_tokens": 19721344.0,
      "step": 10585
    },
    {
      "entropy": 5.936394882202149,
      "epoch": 0.9351819145178382,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004919430277356955,
      "loss": 5.6126,
      "mean_token_accuracy": 0.1858357533812523,
      "num_tokens": 19731105.0,
      "step": 10590
    },
    {
      "entropy": 5.884356546401977,
      "epoch": 0.9356234546096786,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919346738126351,
      "loss": 5.6094,
      "mean_token_accuracy": 0.1931505247950554,
      "num_tokens": 19740162.0,
      "step": 10595
    },
    {
      "entropy": 5.998894453048706,
      "epoch": 0.9360649947015189,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919263156399478,
      "loss": 5.6527,
      "mean_token_accuracy": 0.18412497639656067,
      "num_tokens": 19749228.0,
      "step": 10600
    },
    {
      "entropy": 6.010846328735352,
      "epoch": 0.9365065347933592,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919179532177976,
      "loss": 5.6138,
      "mean_token_accuracy": 0.18535795360803603,
      "num_tokens": 19758642.0,
      "step": 10605
    },
    {
      "entropy": 5.93751072883606,
      "epoch": 0.9369480748851996,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919095865463479,
      "loss": 5.6412,
      "mean_token_accuracy": 0.18951460719108582,
      "num_tokens": 19767923.0,
      "step": 10610
    },
    {
      "entropy": 5.897444581985473,
      "epoch": 0.9373896149770399,
      "grad_norm": 1.0,
      "learning_rate": 0.0004919012156257627,
      "loss": 5.5782,
      "mean_token_accuracy": 0.1900268778204918,
      "num_tokens": 19776204.0,
      "step": 10615
    },
    {
      "entropy": 5.893263435363769,
      "epoch": 0.9378311550688803,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004918928404562061,
      "loss": 5.5787,
      "mean_token_accuracy": 0.19368526488542556,
      "num_tokens": 19785504.0,
      "step": 10620
    },
    {
      "entropy": 5.94614691734314,
      "epoch": 0.9382726951607206,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918844610378421,
      "loss": 5.6208,
      "mean_token_accuracy": 0.18046532571315765,
      "num_tokens": 19794524.0,
      "step": 10625
    },
    {
      "entropy": 5.973563575744629,
      "epoch": 0.9387142352525609,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004918760773708348,
      "loss": 5.6458,
      "mean_token_accuracy": 0.1909143552184105,
      "num_tokens": 19804384.0,
      "step": 10630
    },
    {
      "entropy": 5.936643648147583,
      "epoch": 0.9391557753444013,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004918676894553485,
      "loss": 5.5993,
      "mean_token_accuracy": 0.17671190351247787,
      "num_tokens": 19813893.0,
      "step": 10635
    },
    {
      "entropy": 5.8935009956359865,
      "epoch": 0.9395973154362416,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918592972915476,
      "loss": 5.6604,
      "mean_token_accuracy": 0.18539579063653946,
      "num_tokens": 19823772.0,
      "step": 10640
    },
    {
      "entropy": 5.85434136390686,
      "epoch": 0.940038855528082,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004918509008795963,
      "loss": 5.4664,
      "mean_token_accuracy": 0.1911471500992775,
      "num_tokens": 19832898.0,
      "step": 10645
    },
    {
      "entropy": 6.012303972244263,
      "epoch": 0.9404803956199222,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004918425002196594,
      "loss": 5.6256,
      "mean_token_accuracy": 0.17957846969366073,
      "num_tokens": 19842951.0,
      "step": 10650
    },
    {
      "entropy": 5.987101364135742,
      "epoch": 0.9409219357117626,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491834095311901,
      "loss": 5.6507,
      "mean_token_accuracy": 0.18781149834394456,
      "num_tokens": 19852417.0,
      "step": 10655
    },
    {
      "entropy": 5.984669589996338,
      "epoch": 0.941363475803603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491825686156486,
      "loss": 5.6969,
      "mean_token_accuracy": 0.1760191559791565,
      "num_tokens": 19862377.0,
      "step": 10660
    },
    {
      "entropy": 6.005928707122803,
      "epoch": 0.9418050158954433,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004918172727535792,
      "loss": 5.8424,
      "mean_token_accuracy": 0.18726255297660827,
      "num_tokens": 19871702.0,
      "step": 10665
    },
    {
      "entropy": 5.99537091255188,
      "epoch": 0.9422465559872837,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918088551033454,
      "loss": 5.6483,
      "mean_token_accuracy": 0.18018966764211655,
      "num_tokens": 19881989.0,
      "step": 10670
    },
    {
      "entropy": 5.959221220016479,
      "epoch": 0.942688096079124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004918004332059493,
      "loss": 5.6504,
      "mean_token_accuracy": 0.18416496068239213,
      "num_tokens": 19892130.0,
      "step": 10675
    },
    {
      "entropy": 5.8633129596710205,
      "epoch": 0.9431296361709643,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917920070615561,
      "loss": 5.5428,
      "mean_token_accuracy": 0.19062291383743285,
      "num_tokens": 19901117.0,
      "step": 10680
    },
    {
      "entropy": 5.957319879531861,
      "epoch": 0.9435711762628046,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004917835766703307,
      "loss": 5.6126,
      "mean_token_accuracy": 0.1930697113275528,
      "num_tokens": 19910757.0,
      "step": 10685
    },
    {
      "entropy": 5.953607177734375,
      "epoch": 0.944012716354645,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004917751420324384,
      "loss": 5.608,
      "mean_token_accuracy": 0.18533214330673217,
      "num_tokens": 19920369.0,
      "step": 10690
    },
    {
      "entropy": 5.97283148765564,
      "epoch": 0.9444542564464854,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004917667031480444,
      "loss": 5.6105,
      "mean_token_accuracy": 0.18228392601013182,
      "num_tokens": 19930273.0,
      "step": 10695
    },
    {
      "entropy": 5.919452953338623,
      "epoch": 0.9448957965383257,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917582600173138,
      "loss": 5.5914,
      "mean_token_accuracy": 0.1909669503569603,
      "num_tokens": 19939338.0,
      "step": 10700
    },
    {
      "entropy": 5.941073369979859,
      "epoch": 0.9453373366301661,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917498126404123,
      "loss": 5.6005,
      "mean_token_accuracy": 0.19359002709388734,
      "num_tokens": 19948927.0,
      "step": 10705
    },
    {
      "entropy": 5.875793886184693,
      "epoch": 0.9457788767220063,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917413610175052,
      "loss": 5.558,
      "mean_token_accuracy": 0.19177542626857758,
      "num_tokens": 19958844.0,
      "step": 10710
    },
    {
      "entropy": 5.989517498016357,
      "epoch": 0.9462204168138467,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917329051487582,
      "loss": 5.6415,
      "mean_token_accuracy": 0.18666990250349044,
      "num_tokens": 19968362.0,
      "step": 10715
    },
    {
      "entropy": 5.923425340652466,
      "epoch": 0.946661956905687,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004917244450343368,
      "loss": 5.5832,
      "mean_token_accuracy": 0.19295388460159302,
      "num_tokens": 19977700.0,
      "step": 10720
    },
    {
      "entropy": 5.855211973190308,
      "epoch": 0.9471034969975274,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917159806744067,
      "loss": 5.5107,
      "mean_token_accuracy": 0.19075772762298585,
      "num_tokens": 19986685.0,
      "step": 10725
    },
    {
      "entropy": 5.847012281417847,
      "epoch": 0.9475450370893678,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917075120691338,
      "loss": 5.5536,
      "mean_token_accuracy": 0.1903253376483917,
      "num_tokens": 19997846.0,
      "step": 10730
    },
    {
      "entropy": 5.992723083496093,
      "epoch": 0.947986577181208,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916990392186841,
      "loss": 5.6361,
      "mean_token_accuracy": 0.1826902896165848,
      "num_tokens": 20007530.0,
      "step": 10735
    },
    {
      "entropy": 5.880076742172241,
      "epoch": 0.9484281172730484,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004916905621232233,
      "loss": 5.5272,
      "mean_token_accuracy": 0.19134003967046737,
      "num_tokens": 20017124.0,
      "step": 10740
    },
    {
      "entropy": 5.933245468139648,
      "epoch": 0.9488696573648887,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004916820807829177,
      "loss": 5.6476,
      "mean_token_accuracy": 0.18173979073762894,
      "num_tokens": 20026437.0,
      "step": 10745
    },
    {
      "entropy": 5.888998222351074,
      "epoch": 0.9493111974567291,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916735951979335,
      "loss": 5.5421,
      "mean_token_accuracy": 0.1946476072072983,
      "num_tokens": 20035414.0,
      "step": 10750
    },
    {
      "entropy": 5.82590503692627,
      "epoch": 0.9497527375485694,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004916651053684366,
      "loss": 5.4573,
      "mean_token_accuracy": 0.1965792417526245,
      "num_tokens": 20045508.0,
      "step": 10755
    },
    {
      "entropy": 5.9567255020141605,
      "epoch": 0.9501942776404098,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916566112945936,
      "loss": 5.7017,
      "mean_token_accuracy": 0.18218687772750855,
      "num_tokens": 20055561.0,
      "step": 10760
    },
    {
      "entropy": 5.986366319656372,
      "epoch": 0.9506358177322501,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004916481129765708,
      "loss": 5.6259,
      "mean_token_accuracy": 0.1888494074344635,
      "num_tokens": 20065254.0,
      "step": 10765
    },
    {
      "entropy": 5.892359638214112,
      "epoch": 0.9510773578240904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004916396104145347,
      "loss": 5.5599,
      "mean_token_accuracy": 0.1904727667570114,
      "num_tokens": 20074732.0,
      "step": 10770
    },
    {
      "entropy": 5.859963464736938,
      "epoch": 0.9515188979159308,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004916311036086518,
      "loss": 5.5214,
      "mean_token_accuracy": 0.1960100919008255,
      "num_tokens": 20084921.0,
      "step": 10775
    },
    {
      "entropy": 5.902382230758667,
      "epoch": 0.9519604380077711,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916225925590889,
      "loss": 5.531,
      "mean_token_accuracy": 0.19683946520090104,
      "num_tokens": 20094281.0,
      "step": 10780
    },
    {
      "entropy": 5.874811506271362,
      "epoch": 0.9524019780996115,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916140772660124,
      "loss": 5.5349,
      "mean_token_accuracy": 0.19247579872608184,
      "num_tokens": 20103007.0,
      "step": 10785
    },
    {
      "entropy": 5.938935470581055,
      "epoch": 0.9528435181914517,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916055577295895,
      "loss": 5.6364,
      "mean_token_accuracy": 0.18274036794900894,
      "num_tokens": 20112091.0,
      "step": 10790
    },
    {
      "entropy": 5.905997705459595,
      "epoch": 0.9532850582832921,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915970339499868,
      "loss": 5.5641,
      "mean_token_accuracy": 0.19448008984327317,
      "num_tokens": 20122769.0,
      "step": 10795
    },
    {
      "entropy": 5.8301904678344725,
      "epoch": 0.9537265983751325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915885059273714,
      "loss": 5.6043,
      "mean_token_accuracy": 0.19086600393056868,
      "num_tokens": 20131848.0,
      "step": 10800
    },
    {
      "entropy": 5.985840129852295,
      "epoch": 0.9541681384669728,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915799736619104,
      "loss": 5.6635,
      "mean_token_accuracy": 0.1818881019949913,
      "num_tokens": 20141311.0,
      "step": 10805
    },
    {
      "entropy": 5.98810338973999,
      "epoch": 0.9546096785588132,
      "grad_norm": 1.0,
      "learning_rate": 0.000491571437153771,
      "loss": 5.5726,
      "mean_token_accuracy": 0.18477373719215393,
      "num_tokens": 20150653.0,
      "step": 10810
    },
    {
      "entropy": 5.861916780471802,
      "epoch": 0.9550512186506535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915628964031202,
      "loss": 5.5605,
      "mean_token_accuracy": 0.19274954944849015,
      "num_tokens": 20160068.0,
      "step": 10815
    },
    {
      "entropy": 5.9410929679870605,
      "epoch": 0.9554927587424938,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004915543514101253,
      "loss": 5.6383,
      "mean_token_accuracy": 0.18524001389741898,
      "num_tokens": 20170309.0,
      "step": 10820
    },
    {
      "entropy": 5.882420349121094,
      "epoch": 0.9559342988343341,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004915458021749541,
      "loss": 5.5534,
      "mean_token_accuracy": 0.19386060237884523,
      "num_tokens": 20180037.0,
      "step": 10825
    },
    {
      "entropy": 5.97072548866272,
      "epoch": 0.9563758389261745,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915372486977736,
      "loss": 5.6104,
      "mean_token_accuracy": 0.18328188806772233,
      "num_tokens": 20189134.0,
      "step": 10830
    },
    {
      "entropy": 5.9443323612213135,
      "epoch": 0.9568173790180149,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915286909787516,
      "loss": 5.57,
      "mean_token_accuracy": 0.18475655913352967,
      "num_tokens": 20199295.0,
      "step": 10835
    },
    {
      "entropy": 5.811439895629883,
      "epoch": 0.9572589191098552,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004915201290180556,
      "loss": 5.5546,
      "mean_token_accuracy": 0.19458646029233934,
      "num_tokens": 20207780.0,
      "step": 10840
    },
    {
      "entropy": 5.874908351898194,
      "epoch": 0.9577004592016956,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004915115628158535,
      "loss": 5.4554,
      "mean_token_accuracy": 0.1929043412208557,
      "num_tokens": 20216743.0,
      "step": 10845
    },
    {
      "entropy": 5.892486715316773,
      "epoch": 0.9581419992935358,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491502992372313,
      "loss": 5.5279,
      "mean_token_accuracy": 0.19214529544115067,
      "num_tokens": 20225539.0,
      "step": 10850
    },
    {
      "entropy": 5.858796310424805,
      "epoch": 0.9585835393853762,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491494417687602,
      "loss": 5.5218,
      "mean_token_accuracy": 0.1890709638595581,
      "num_tokens": 20234613.0,
      "step": 10855
    },
    {
      "entropy": 5.929288005828857,
      "epoch": 0.9590250794772165,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004914858387618886,
      "loss": 5.6148,
      "mean_token_accuracy": 0.19103958308696747,
      "num_tokens": 20243681.0,
      "step": 10860
    },
    {
      "entropy": 5.8858997344970705,
      "epoch": 0.9594666195690569,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004914772555953406,
      "loss": 5.5209,
      "mean_token_accuracy": 0.19760328829288482,
      "num_tokens": 20253209.0,
      "step": 10865
    },
    {
      "entropy": 5.869548559188843,
      "epoch": 0.9599081596608973,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004914686681881264,
      "loss": 5.6002,
      "mean_token_accuracy": 0.19028532803058623,
      "num_tokens": 20263670.0,
      "step": 10870
    },
    {
      "entropy": 5.988346195220947,
      "epoch": 0.9603496997527375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000491460076540414,
      "loss": 5.6065,
      "mean_token_accuracy": 0.1852234721183777,
      "num_tokens": 20272106.0,
      "step": 10875
    },
    {
      "entropy": 5.951537656784057,
      "epoch": 0.9607912398445779,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914514806523718,
      "loss": 5.5472,
      "mean_token_accuracy": 0.19902791380882262,
      "num_tokens": 20280530.0,
      "step": 10880
    },
    {
      "entropy": 5.887434720993042,
      "epoch": 0.9612327799364182,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004914428805241683,
      "loss": 5.5532,
      "mean_token_accuracy": 0.18996213078498841,
      "num_tokens": 20289347.0,
      "step": 10885
    },
    {
      "entropy": 5.808625268936157,
      "epoch": 0.9616743200282586,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004914342761559718,
      "loss": 5.5395,
      "mean_token_accuracy": 0.1942080870270729,
      "num_tokens": 20298350.0,
      "step": 10890
    },
    {
      "entropy": 5.962233638763427,
      "epoch": 0.9621158601200989,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914256675479509,
      "loss": 5.6564,
      "mean_token_accuracy": 0.1755989134311676,
      "num_tokens": 20308028.0,
      "step": 10895
    },
    {
      "entropy": 5.882785940170288,
      "epoch": 0.9625574002119393,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004914170547002742,
      "loss": 5.5681,
      "mean_token_accuracy": 0.19408486187458038,
      "num_tokens": 20316878.0,
      "step": 10900
    },
    {
      "entropy": 5.868825197219849,
      "epoch": 0.9629989403037796,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914084376131106,
      "loss": 5.4861,
      "mean_token_accuracy": 0.19691345989704132,
      "num_tokens": 20326272.0,
      "step": 10905
    },
    {
      "entropy": 5.905012512207032,
      "epoch": 0.9634404803956199,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004913998162866288,
      "loss": 5.4335,
      "mean_token_accuracy": 0.205233034491539,
      "num_tokens": 20334274.0,
      "step": 10910
    },
    {
      "entropy": 5.924832725524903,
      "epoch": 0.9638820204874603,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004913911907209976,
      "loss": 5.6502,
      "mean_token_accuracy": 0.1869783580303192,
      "num_tokens": 20344126.0,
      "step": 10915
    },
    {
      "entropy": 5.8474640369415285,
      "epoch": 0.9643235605793006,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004913825609163861,
      "loss": 5.5232,
      "mean_token_accuracy": 0.19692277163267136,
      "num_tokens": 20352975.0,
      "step": 10920
    },
    {
      "entropy": 5.984229183197021,
      "epoch": 0.964765100671141,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004913739268729633,
      "loss": 5.5995,
      "mean_token_accuracy": 0.1848618745803833,
      "num_tokens": 20362141.0,
      "step": 10925
    },
    {
      "entropy": 5.805004119873047,
      "epoch": 0.9652066407629812,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004913652885908983,
      "loss": 5.5673,
      "mean_token_accuracy": 0.18914556503295898,
      "num_tokens": 20371827.0,
      "step": 10930
    },
    {
      "entropy": 5.867023563385009,
      "epoch": 0.9656481808548216,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004913566460703603,
      "loss": 5.5755,
      "mean_token_accuracy": 0.19183651059865953,
      "num_tokens": 20380885.0,
      "step": 10935
    },
    {
      "entropy": 5.951615905761718,
      "epoch": 0.966089720946662,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004913479993115187,
      "loss": 5.5628,
      "mean_token_accuracy": 0.1906025692820549,
      "num_tokens": 20390142.0,
      "step": 10940
    },
    {
      "entropy": 5.919005680084228,
      "epoch": 0.9665312610385023,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004913393483145428,
      "loss": 5.6033,
      "mean_token_accuracy": 0.19135210812091827,
      "num_tokens": 20401162.0,
      "step": 10945
    },
    {
      "entropy": 5.816787433624268,
      "epoch": 0.9669728011303427,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004913306930796022,
      "loss": 5.5491,
      "mean_token_accuracy": 0.18465678095817567,
      "num_tokens": 20410831.0,
      "step": 10950
    },
    {
      "entropy": 5.964951992034912,
      "epoch": 0.967414341222183,
      "grad_norm": 0.875,
      "learning_rate": 0.0004913220336068662,
      "loss": 5.545,
      "mean_token_accuracy": 0.1931534692645073,
      "num_tokens": 20419822.0,
      "step": 10955
    },
    {
      "entropy": 5.893057060241699,
      "epoch": 0.9678558813140233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913133698965048,
      "loss": 5.5181,
      "mean_token_accuracy": 0.19152877032756804,
      "num_tokens": 20429517.0,
      "step": 10960
    },
    {
      "entropy": 5.870708036422729,
      "epoch": 0.9682974214058636,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004913047019486874,
      "loss": 5.6181,
      "mean_token_accuracy": 0.1859910562634468,
      "num_tokens": 20438997.0,
      "step": 10965
    },
    {
      "entropy": 5.9326488971710205,
      "epoch": 0.968738961497704,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004912960297635839,
      "loss": 5.6278,
      "mean_token_accuracy": 0.1850393980741501,
      "num_tokens": 20448799.0,
      "step": 10970
    },
    {
      "entropy": 5.901592350006103,
      "epoch": 0.9691805015895444,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004912873533413641,
      "loss": 5.5579,
      "mean_token_accuracy": 0.1908988282084465,
      "num_tokens": 20458347.0,
      "step": 10975
    },
    {
      "entropy": 5.858342790603638,
      "epoch": 0.9696220416813847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912786726821982,
      "loss": 5.5253,
      "mean_token_accuracy": 0.19447053223848343,
      "num_tokens": 20467223.0,
      "step": 10980
    },
    {
      "entropy": 5.907288932800293,
      "epoch": 0.970063581773225,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491269987786256,
      "loss": 5.5717,
      "mean_token_accuracy": 0.19068745970726014,
      "num_tokens": 20477103.0,
      "step": 10985
    },
    {
      "entropy": 5.988813734054565,
      "epoch": 0.9705051218650653,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004912612986537078,
      "loss": 5.6176,
      "mean_token_accuracy": 0.18468385934829712,
      "num_tokens": 20487691.0,
      "step": 10990
    },
    {
      "entropy": 5.930143928527832,
      "epoch": 0.9709466619569057,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004912526052847238,
      "loss": 5.5513,
      "mean_token_accuracy": 0.19809958934783936,
      "num_tokens": 20496106.0,
      "step": 10995
    },
    {
      "entropy": 5.869368076324463,
      "epoch": 0.971388202048746,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004912439076794742,
      "loss": 5.4744,
      "mean_token_accuracy": 0.20135725140571595,
      "num_tokens": 20506126.0,
      "step": 11000
    },
    {
      "entropy": 5.884821224212646,
      "epoch": 0.9718297421405864,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004912352058381293,
      "loss": 5.62,
      "mean_token_accuracy": 0.18596713840961457,
      "num_tokens": 20515242.0,
      "step": 11005
    },
    {
      "entropy": 5.925478219985962,
      "epoch": 0.9722712822324268,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004912264997608599,
      "loss": 5.552,
      "mean_token_accuracy": 0.19096482396125794,
      "num_tokens": 20524320.0,
      "step": 11010
    },
    {
      "entropy": 5.807953834533691,
      "epoch": 0.972712822324267,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004912177894478364,
      "loss": 5.437,
      "mean_token_accuracy": 0.20357599556446077,
      "num_tokens": 20533450.0,
      "step": 11015
    },
    {
      "entropy": 5.854403257369995,
      "epoch": 0.9731543624161074,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004912090748992291,
      "loss": 5.4495,
      "mean_token_accuracy": 0.1984184965491295,
      "num_tokens": 20542922.0,
      "step": 11020
    },
    {
      "entropy": 5.902002000808716,
      "epoch": 0.9735959025079477,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912003561152092,
      "loss": 5.4885,
      "mean_token_accuracy": 0.20224822759628297,
      "num_tokens": 20551523.0,
      "step": 11025
    },
    {
      "entropy": 5.805849409103393,
      "epoch": 0.9740374425997881,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911916330959472,
      "loss": 5.4738,
      "mean_token_accuracy": 0.19958919882774354,
      "num_tokens": 20561765.0,
      "step": 11030
    },
    {
      "entropy": 5.851793193817139,
      "epoch": 0.9744789826916284,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004911829058416141,
      "loss": 5.5409,
      "mean_token_accuracy": 0.18966053128242494,
      "num_tokens": 20571028.0,
      "step": 11035
    },
    {
      "entropy": 5.955337905883789,
      "epoch": 0.9749205227834687,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911741743523808,
      "loss": 5.6235,
      "mean_token_accuracy": 0.18315792828798294,
      "num_tokens": 20579992.0,
      "step": 11040
    },
    {
      "entropy": 5.892729616165161,
      "epoch": 0.9753620628753091,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004911654386284184,
      "loss": 5.5242,
      "mean_token_accuracy": 0.19104957729578018,
      "num_tokens": 20588860.0,
      "step": 11045
    },
    {
      "entropy": 5.919535112380982,
      "epoch": 0.9758036029671494,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000491156698669898,
      "loss": 5.5928,
      "mean_token_accuracy": 0.18698275536298753,
      "num_tokens": 20598210.0,
      "step": 11050
    },
    {
      "entropy": 5.915886116027832,
      "epoch": 0.9762451430589898,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911479544769908,
      "loss": 5.5653,
      "mean_token_accuracy": 0.19796052128076552,
      "num_tokens": 20607058.0,
      "step": 11055
    },
    {
      "entropy": 5.881275320053101,
      "epoch": 0.9766866831508301,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491139206049868,
      "loss": 5.5635,
      "mean_token_accuracy": 0.19711262881755828,
      "num_tokens": 20616509.0,
      "step": 11060
    },
    {
      "entropy": 5.890599012374878,
      "epoch": 0.9771282232426705,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911304533887011,
      "loss": 5.5027,
      "mean_token_accuracy": 0.19427365362644194,
      "num_tokens": 20625710.0,
      "step": 11065
    },
    {
      "entropy": 5.9025853157043455,
      "epoch": 0.9775697633345107,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911216964936616,
      "loss": 5.5354,
      "mean_token_accuracy": 0.19517610371112823,
      "num_tokens": 20634230.0,
      "step": 11070
    },
    {
      "entropy": 5.912288427352905,
      "epoch": 0.9780113034263511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004911129353649211,
      "loss": 5.5536,
      "mean_token_accuracy": 0.19475232809782028,
      "num_tokens": 20643349.0,
      "step": 11075
    },
    {
      "entropy": 5.913419437408447,
      "epoch": 0.9784528435181915,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000491104170002651,
      "loss": 5.5343,
      "mean_token_accuracy": 0.19227563589811325,
      "num_tokens": 20653168.0,
      "step": 11080
    },
    {
      "entropy": 5.908128356933593,
      "epoch": 0.9788943836100318,
      "grad_norm": 1.0625,
      "learning_rate": 0.000491095400407023,
      "loss": 5.5975,
      "mean_token_accuracy": 0.1934755951166153,
      "num_tokens": 20663829.0,
      "step": 11085
    },
    {
      "entropy": 5.837201929092407,
      "epoch": 0.9793359237018722,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004910866265782091,
      "loss": 5.546,
      "mean_token_accuracy": 0.1963401034474373,
      "num_tokens": 20672603.0,
      "step": 11090
    },
    {
      "entropy": 5.909215879440308,
      "epoch": 0.9797774637937124,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004910778485163811,
      "loss": 5.5172,
      "mean_token_accuracy": 0.19357363134622574,
      "num_tokens": 20681358.0,
      "step": 11095
    },
    {
      "entropy": 5.862726926803589,
      "epoch": 0.9802190038855528,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000491069066221711,
      "loss": 5.581,
      "mean_token_accuracy": 0.18739733397960662,
      "num_tokens": 20691557.0,
      "step": 11100
    },
    {
      "entropy": 5.876424837112427,
      "epoch": 0.9806605439773931,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004910602796943707,
      "loss": 5.5657,
      "mean_token_accuracy": 0.18983499854803085,
      "num_tokens": 20701057.0,
      "step": 11105
    },
    {
      "entropy": 5.833113336563111,
      "epoch": 0.9811020840692335,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004910514889345323,
      "loss": 5.5239,
      "mean_token_accuracy": 0.1907399535179138,
      "num_tokens": 20709725.0,
      "step": 11110
    },
    {
      "entropy": 5.94479603767395,
      "epoch": 0.9815436241610739,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910426939423682,
      "loss": 5.5865,
      "mean_token_accuracy": 0.19195503890514373,
      "num_tokens": 20718929.0,
      "step": 11115
    },
    {
      "entropy": 5.887921476364136,
      "epoch": 0.9819851642529142,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004910338947180506,
      "loss": 5.5759,
      "mean_token_accuracy": 0.18831754028797149,
      "num_tokens": 20728401.0,
      "step": 11120
    },
    {
      "entropy": 5.905521011352539,
      "epoch": 0.9824267043447545,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004910250912617519,
      "loss": 5.6401,
      "mean_token_accuracy": 0.1841378778219223,
      "num_tokens": 20737107.0,
      "step": 11125
    },
    {
      "entropy": 5.982414579391479,
      "epoch": 0.9828682444365948,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004910162835736445,
      "loss": 5.637,
      "mean_token_accuracy": 0.18379368931055068,
      "num_tokens": 20747095.0,
      "step": 11130
    },
    {
      "entropy": 5.933509826660156,
      "epoch": 0.9833097845284352,
      "grad_norm": 0.984375,
      "learning_rate": 0.000491007471653901,
      "loss": 5.5764,
      "mean_token_accuracy": 0.1879402682185173,
      "num_tokens": 20756425.0,
      "step": 11135
    },
    {
      "entropy": 5.931452417373658,
      "epoch": 0.9837513246202755,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909986555026939,
      "loss": 5.5485,
      "mean_token_accuracy": 0.19237802624702455,
      "num_tokens": 20764823.0,
      "step": 11140
    },
    {
      "entropy": 5.8991471290588375,
      "epoch": 0.9841928647121159,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490989835120196,
      "loss": 5.6028,
      "mean_token_accuracy": 0.1867647334933281,
      "num_tokens": 20775360.0,
      "step": 11145
    },
    {
      "entropy": 5.910455942153931,
      "epoch": 0.9846344048039563,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909810105065801,
      "loss": 5.6308,
      "mean_token_accuracy": 0.18141679018735885,
      "num_tokens": 20784389.0,
      "step": 11150
    },
    {
      "entropy": 5.913553380966187,
      "epoch": 0.9850759448957965,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004909721816620191,
      "loss": 5.5626,
      "mean_token_accuracy": 0.18852440118789673,
      "num_tokens": 20793581.0,
      "step": 11155
    },
    {
      "entropy": 6.005950212478638,
      "epoch": 0.9855174849876369,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909633485866857,
      "loss": 5.6139,
      "mean_token_accuracy": 0.19549831748008728,
      "num_tokens": 20803409.0,
      "step": 11160
    },
    {
      "entropy": 5.8698395729064945,
      "epoch": 0.9859590250794772,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004909545112807533,
      "loss": 5.5933,
      "mean_token_accuracy": 0.19143689125776292,
      "num_tokens": 20812630.0,
      "step": 11165
    },
    {
      "entropy": 5.883950805664062,
      "epoch": 0.9864005651713176,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004909456697443948,
      "loss": 5.5845,
      "mean_token_accuracy": 0.1920734167098999,
      "num_tokens": 20821682.0,
      "step": 11170
    },
    {
      "entropy": 5.909452962875366,
      "epoch": 0.9868421052631579,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004909368239777835,
      "loss": 5.5986,
      "mean_token_accuracy": 0.19298047721385955,
      "num_tokens": 20830882.0,
      "step": 11175
    },
    {
      "entropy": 5.921415185928344,
      "epoch": 0.9872836453549982,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004909279739810926,
      "loss": 5.5425,
      "mean_token_accuracy": 0.18241304755210877,
      "num_tokens": 20839841.0,
      "step": 11180
    },
    {
      "entropy": 5.868939924240112,
      "epoch": 0.9877251854468386,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909191197544954,
      "loss": 5.568,
      "mean_token_accuracy": 0.19773616045713424,
      "num_tokens": 20849764.0,
      "step": 11185
    },
    {
      "entropy": 5.881763172149658,
      "epoch": 0.9881667255386789,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004909102612981655,
      "loss": 5.669,
      "mean_token_accuracy": 0.1827738791704178,
      "num_tokens": 20859061.0,
      "step": 11190
    },
    {
      "entropy": 5.937941265106201,
      "epoch": 0.9886082656305193,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004909013986122763,
      "loss": 5.5433,
      "mean_token_accuracy": 0.19337455779314042,
      "num_tokens": 20868952.0,
      "step": 11195
    },
    {
      "entropy": 5.954920148849487,
      "epoch": 0.9890498057223596,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004908925316970016,
      "loss": 5.5249,
      "mean_token_accuracy": 0.19236643612384796,
      "num_tokens": 20879050.0,
      "step": 11200
    },
    {
      "entropy": 5.93258056640625,
      "epoch": 0.9894913458142,
      "grad_norm": 0.875,
      "learning_rate": 0.000490883660552515,
      "loss": 5.5928,
      "mean_token_accuracy": 0.19040617197752,
      "num_tokens": 20888849.0,
      "step": 11205
    },
    {
      "entropy": 5.910648155212402,
      "epoch": 0.9899328859060402,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908747851789902,
      "loss": 5.548,
      "mean_token_accuracy": 0.1922212466597557,
      "num_tokens": 20897551.0,
      "step": 11210
    },
    {
      "entropy": 5.956227970123291,
      "epoch": 0.9903744259978806,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004908659055766011,
      "loss": 5.6653,
      "mean_token_accuracy": 0.1858624890446663,
      "num_tokens": 20907670.0,
      "step": 11215
    },
    {
      "entropy": 5.9300590515136715,
      "epoch": 0.990815966089721,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004908570217455218,
      "loss": 5.5352,
      "mean_token_accuracy": 0.18535751551389695,
      "num_tokens": 20917050.0,
      "step": 11220
    },
    {
      "entropy": 5.89700288772583,
      "epoch": 0.9912575061815613,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004908481336859262,
      "loss": 5.5498,
      "mean_token_accuracy": 0.1924524948000908,
      "num_tokens": 20925747.0,
      "step": 11225
    },
    {
      "entropy": 5.901921987533569,
      "epoch": 0.9916990462734017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908392413979884,
      "loss": 5.5429,
      "mean_token_accuracy": 0.1929529696702957,
      "num_tokens": 20934367.0,
      "step": 11230
    },
    {
      "entropy": 5.823864889144898,
      "epoch": 0.9921405863652419,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004908303448818825,
      "loss": 5.459,
      "mean_token_accuracy": 0.1900830402970314,
      "num_tokens": 20943700.0,
      "step": 11235
    },
    {
      "entropy": 5.905262613296509,
      "epoch": 0.9925821264570823,
      "grad_norm": 0.9375,
      "learning_rate": 0.000490821444137783,
      "loss": 5.5963,
      "mean_token_accuracy": 0.19443309903144837,
      "num_tokens": 20953043.0,
      "step": 11240
    },
    {
      "entropy": 5.9580878734588625,
      "epoch": 0.9930236665489226,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004908125391658641,
      "loss": 5.6323,
      "mean_token_accuracy": 0.18162959814071655,
      "num_tokens": 20962036.0,
      "step": 11245
    },
    {
      "entropy": 5.934054517745972,
      "epoch": 0.993465206640763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908036299663003,
      "loss": 5.5639,
      "mean_token_accuracy": 0.18985943496227264,
      "num_tokens": 20971831.0,
      "step": 11250
    },
    {
      "entropy": 5.903661155700684,
      "epoch": 0.9939067467326034,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004907947165392662,
      "loss": 5.6368,
      "mean_token_accuracy": 0.18064841777086257,
      "num_tokens": 20981787.0,
      "step": 11255
    },
    {
      "entropy": 5.899556398391724,
      "epoch": 0.9943482868244437,
      "grad_norm": 1.0,
      "learning_rate": 0.0004907857988849363,
      "loss": 5.4827,
      "mean_token_accuracy": 0.19780028164386748,
      "num_tokens": 20991475.0,
      "step": 11260
    },
    {
      "entropy": 5.911313247680664,
      "epoch": 0.994789826916284,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004907768770034853,
      "loss": 5.5522,
      "mean_token_accuracy": 0.19119220077991486,
      "num_tokens": 21000756.0,
      "step": 11265
    },
    {
      "entropy": 5.835139179229737,
      "epoch": 0.9952313670081243,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000490767950895088,
      "loss": 5.5753,
      "mean_token_accuracy": 0.19002063423395157,
      "num_tokens": 21010573.0,
      "step": 11270
    },
    {
      "entropy": 5.878307676315307,
      "epoch": 0.9956729070999647,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004907590205599192,
      "loss": 5.5835,
      "mean_token_accuracy": 0.2029650777578354,
      "num_tokens": 21020003.0,
      "step": 11275
    },
    {
      "entropy": 5.869453239440918,
      "epoch": 0.996114447191805,
      "grad_norm": 0.953125,
      "learning_rate": 0.000490750085998154,
      "loss": 5.5346,
      "mean_token_accuracy": 0.18952150344848634,
      "num_tokens": 21028627.0,
      "step": 11280
    },
    {
      "entropy": 5.894145822525024,
      "epoch": 0.9965559872836454,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004907411472099672,
      "loss": 5.5671,
      "mean_token_accuracy": 0.19857637137174605,
      "num_tokens": 21037672.0,
      "step": 11285
    },
    {
      "entropy": 5.918149852752686,
      "epoch": 0.9969975273754857,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004907322041955341,
      "loss": 5.6515,
      "mean_token_accuracy": 0.182760551571846,
      "num_tokens": 21047130.0,
      "step": 11290
    },
    {
      "entropy": 5.995540380477905,
      "epoch": 0.997439067467326,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004907232569550298,
      "loss": 5.6032,
      "mean_token_accuracy": 0.19152709394693374,
      "num_tokens": 21057016.0,
      "step": 11295
    },
    {
      "entropy": 5.916431713104248,
      "epoch": 0.9978806075591664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004907143054886297,
      "loss": 5.6363,
      "mean_token_accuracy": 0.18286186307668686,
      "num_tokens": 21066275.0,
      "step": 11300
    },
    {
      "entropy": 5.881349420547485,
      "epoch": 0.9983221476510067,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000490705349796509,
      "loss": 5.6592,
      "mean_token_accuracy": 0.1859086975455284,
      "num_tokens": 21077098.0,
      "step": 11305
    },
    {
      "entropy": 5.956093120574951,
      "epoch": 0.9987636877428471,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004906963898788431,
      "loss": 5.6093,
      "mean_token_accuracy": 0.18696858137845992,
      "num_tokens": 21087041.0,
      "step": 11310
    },
    {
      "entropy": 5.947420072555542,
      "epoch": 0.9992052278346873,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004906874257358075,
      "loss": 5.5178,
      "mean_token_accuracy": 0.18915889114141465,
      "num_tokens": 21096450.0,
      "step": 11315
    },
    {
      "entropy": 5.861836194992065,
      "epoch": 0.9996467679265277,
      "grad_norm": 0.984375,
      "learning_rate": 0.000490678457367578,
      "loss": 5.5304,
      "mean_token_accuracy": 0.19978671073913573,
      "num_tokens": 21106196.0,
      "step": 11320
    },
    {
      "entropy": 5.770443916320801,
      "epoch": 1.000088308018368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004906694847743302,
      "loss": 5.4564,
      "mean_token_accuracy": 0.20402325689792633,
      "num_tokens": 21114660.0,
      "step": 11325
    },
    {
      "entropy": 5.9286048889160154,
      "epoch": 1.0005298481102085,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004906605079562399,
      "loss": 5.5442,
      "mean_token_accuracy": 0.18752799779176713,
      "num_tokens": 21123885.0,
      "step": 11330
    },
    {
      "entropy": 5.900831413269043,
      "epoch": 1.0009713882020488,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004906515269134827,
      "loss": 5.4911,
      "mean_token_accuracy": 0.19872518628835678,
      "num_tokens": 21132774.0,
      "step": 11335
    },
    {
      "entropy": 5.868030214309693,
      "epoch": 1.001412928293889,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004906425416462349,
      "loss": 5.5324,
      "mean_token_accuracy": 0.1903938338160515,
      "num_tokens": 21141777.0,
      "step": 11340
    },
    {
      "entropy": 5.969753408432007,
      "epoch": 1.0018544683857293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004906335521546723,
      "loss": 5.5613,
      "mean_token_accuracy": 0.18468785285949707,
      "num_tokens": 21151065.0,
      "step": 11345
    },
    {
      "entropy": 5.939084815979004,
      "epoch": 1.0022960084775698,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906245584389712,
      "loss": 5.4608,
      "mean_token_accuracy": 0.1919092580676079,
      "num_tokens": 21160247.0,
      "step": 11350
    },
    {
      "entropy": 5.900304651260376,
      "epoch": 1.00273754856941,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004906155604993075,
      "loss": 5.4959,
      "mean_token_accuracy": 0.1908602997660637,
      "num_tokens": 21169753.0,
      "step": 11355
    },
    {
      "entropy": 5.883688735961914,
      "epoch": 1.0031790886612504,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004906065583358576,
      "loss": 5.4872,
      "mean_token_accuracy": 0.19097884446382524,
      "num_tokens": 21178412.0,
      "step": 11360
    },
    {
      "entropy": 5.904987573623657,
      "epoch": 1.0036206287530909,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905975519487978,
      "loss": 5.5189,
      "mean_token_accuracy": 0.19409771263599396,
      "num_tokens": 21187886.0,
      "step": 11365
    },
    {
      "entropy": 5.875783634185791,
      "epoch": 1.0040621688449312,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004905885413383046,
      "loss": 5.4895,
      "mean_token_accuracy": 0.19214787781238557,
      "num_tokens": 21197450.0,
      "step": 11370
    },
    {
      "entropy": 5.894489479064942,
      "epoch": 1.0045037089367714,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905795265045546,
      "loss": 5.4711,
      "mean_token_accuracy": 0.19499551504850388,
      "num_tokens": 21206989.0,
      "step": 11375
    },
    {
      "entropy": 5.897228002548218,
      "epoch": 1.0049452490286117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004905705074477242,
      "loss": 5.4482,
      "mean_token_accuracy": 0.19684029817581178,
      "num_tokens": 21215543.0,
      "step": 11380
    },
    {
      "entropy": 5.917811441421509,
      "epoch": 1.0053867891204522,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004905614841679901,
      "loss": 5.5605,
      "mean_token_accuracy": 0.1911270335316658,
      "num_tokens": 21224916.0,
      "step": 11385
    },
    {
      "entropy": 5.820360469818115,
      "epoch": 1.0058283292122925,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004905524566655292,
      "loss": 5.3622,
      "mean_token_accuracy": 0.19745410978794098,
      "num_tokens": 21233323.0,
      "step": 11390
    },
    {
      "entropy": 5.876035451889038,
      "epoch": 1.0062698693041328,
      "grad_norm": 1.0,
      "learning_rate": 0.0004905434249405183,
      "loss": 5.4275,
      "mean_token_accuracy": 0.19987936019897462,
      "num_tokens": 21242238.0,
      "step": 11395
    },
    {
      "entropy": 5.875967311859131,
      "epoch": 1.0067114093959733,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004905343889931342,
      "loss": 5.5351,
      "mean_token_accuracy": 0.18856841772794725,
      "num_tokens": 21252621.0,
      "step": 11400
    },
    {
      "entropy": 5.890102910995483,
      "epoch": 1.0071529494878135,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000490525348823554,
      "loss": 5.4387,
      "mean_token_accuracy": 0.1999368414282799,
      "num_tokens": 21261488.0,
      "step": 11405
    },
    {
      "entropy": 5.922180318832398,
      "epoch": 1.0075944895796538,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004905163044319549,
      "loss": 5.4232,
      "mean_token_accuracy": 0.19712914824485778,
      "num_tokens": 21270219.0,
      "step": 11410
    },
    {
      "entropy": 5.877050733566284,
      "epoch": 1.008036029671494,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004905072558185139,
      "loss": 5.4656,
      "mean_token_accuracy": 0.20057197660207748,
      "num_tokens": 21279778.0,
      "step": 11415
    },
    {
      "entropy": 5.909356641769409,
      "epoch": 1.0084775697633346,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004904982029834083,
      "loss": 5.4758,
      "mean_token_accuracy": 0.18741813600063323,
      "num_tokens": 21288608.0,
      "step": 11420
    },
    {
      "entropy": 5.828989696502686,
      "epoch": 1.0089191098551749,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004904891459268154,
      "loss": 5.4941,
      "mean_token_accuracy": 0.1923238903284073,
      "num_tokens": 21298587.0,
      "step": 11425
    },
    {
      "entropy": 5.890131950378418,
      "epoch": 1.0093606499470151,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004904800846489128,
      "loss": 5.4584,
      "mean_token_accuracy": 0.19254144579172133,
      "num_tokens": 21307841.0,
      "step": 11430
    },
    {
      "entropy": 5.921647787094116,
      "epoch": 1.0098021900388556,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004904710191498779,
      "loss": 5.4866,
      "mean_token_accuracy": 0.18880448639392852,
      "num_tokens": 21317299.0,
      "step": 11435
    },
    {
      "entropy": 5.878640937805176,
      "epoch": 1.010243730130696,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004904619494298882,
      "loss": 5.4974,
      "mean_token_accuracy": 0.1968512237071991,
      "num_tokens": 21327427.0,
      "step": 11440
    },
    {
      "entropy": 5.929252433776855,
      "epoch": 1.0106852702225362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004904528754891215,
      "loss": 5.5119,
      "mean_token_accuracy": 0.19123229533433914,
      "num_tokens": 21337126.0,
      "step": 11445
    },
    {
      "entropy": 5.871789455413818,
      "epoch": 1.0111268103143765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904437973277555,
      "loss": 5.4432,
      "mean_token_accuracy": 0.19388150572776794,
      "num_tokens": 21346549.0,
      "step": 11450
    },
    {
      "entropy": 5.827709817886353,
      "epoch": 1.011568350406217,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904347149459681,
      "loss": 5.4464,
      "mean_token_accuracy": 0.19859617054462433,
      "num_tokens": 21355753.0,
      "step": 11455
    },
    {
      "entropy": 5.889517211914063,
      "epoch": 1.0120098904980572,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004904256283439371,
      "loss": 5.482,
      "mean_token_accuracy": 0.19921454638242722,
      "num_tokens": 21365924.0,
      "step": 11460
    },
    {
      "entropy": 5.910828351974487,
      "epoch": 1.0124514305898975,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904165375218405,
      "loss": 5.4374,
      "mean_token_accuracy": 0.20360911041498184,
      "num_tokens": 21375307.0,
      "step": 11465
    },
    {
      "entropy": 5.907249307632446,
      "epoch": 1.012892970681738,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004904074424798565,
      "loss": 5.4847,
      "mean_token_accuracy": 0.19290790259838103,
      "num_tokens": 21384667.0,
      "step": 11470
    },
    {
      "entropy": 5.883868217468262,
      "epoch": 1.0133345107735783,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004903983432181632,
      "loss": 5.5552,
      "mean_token_accuracy": 0.18648406565189363,
      "num_tokens": 21393806.0,
      "step": 11475
    },
    {
      "entropy": 5.955294942855835,
      "epoch": 1.0137760508654186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004903892397369388,
      "loss": 5.4613,
      "mean_token_accuracy": 0.19508447349071503,
      "num_tokens": 21404015.0,
      "step": 11480
    },
    {
      "entropy": 5.827986574172973,
      "epoch": 1.0142175909572588,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004903801320363618,
      "loss": 5.4612,
      "mean_token_accuracy": 0.19046614915132523,
      "num_tokens": 21413581.0,
      "step": 11485
    },
    {
      "entropy": 5.797435760498047,
      "epoch": 1.0146591310490993,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903710201166105,
      "loss": 5.5171,
      "mean_token_accuracy": 0.18126242309808732,
      "num_tokens": 21422178.0,
      "step": 11490
    },
    {
      "entropy": 5.8487104892730715,
      "epoch": 1.0151006711409396,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004903619039778634,
      "loss": 5.4323,
      "mean_token_accuracy": 0.2022449791431427,
      "num_tokens": 21432188.0,
      "step": 11495
    },
    {
      "entropy": 5.929657649993897,
      "epoch": 1.0155422112327799,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000490352783620299,
      "loss": 5.3973,
      "mean_token_accuracy": 0.2076668456196785,
      "num_tokens": 21441172.0,
      "step": 11500
    },
    {
      "entropy": 5.874224281311035,
      "epoch": 1.0159837513246204,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903436590440961,
      "loss": 5.5737,
      "mean_token_accuracy": 0.18894722163677216,
      "num_tokens": 21451211.0,
      "step": 11505
    },
    {
      "entropy": 5.796626806259155,
      "epoch": 1.0164252914164607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004903345302494334,
      "loss": 5.4354,
      "mean_token_accuracy": 0.20208420008420944,
      "num_tokens": 21460446.0,
      "step": 11510
    },
    {
      "entropy": 5.946972036361695,
      "epoch": 1.016866831508301,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004903253972364897,
      "loss": 5.5448,
      "mean_token_accuracy": 0.19134037494659423,
      "num_tokens": 21470690.0,
      "step": 11515
    },
    {
      "entropy": 5.946447658538818,
      "epoch": 1.0173083716001412,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903162600054439,
      "loss": 5.4861,
      "mean_token_accuracy": 0.195264632999897,
      "num_tokens": 21479712.0,
      "step": 11520
    },
    {
      "entropy": 5.891729497909546,
      "epoch": 1.0177499116919817,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903071185564751,
      "loss": 5.4948,
      "mean_token_accuracy": 0.19007141143083572,
      "num_tokens": 21489057.0,
      "step": 11525
    },
    {
      "entropy": 5.835257148742675,
      "epoch": 1.018191451783822,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004902979728897623,
      "loss": 5.4798,
      "mean_token_accuracy": 0.19294410198926926,
      "num_tokens": 21498625.0,
      "step": 11530
    },
    {
      "entropy": 5.968158864974976,
      "epoch": 1.0186329918756623,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004902888230054846,
      "loss": 5.5499,
      "mean_token_accuracy": 0.18832134902477266,
      "num_tokens": 21507918.0,
      "step": 11535
    },
    {
      "entropy": 5.9240625381469725,
      "epoch": 1.0190745319675028,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004902796689038213,
      "loss": 5.4219,
      "mean_token_accuracy": 0.20519990772008895,
      "num_tokens": 21517131.0,
      "step": 11540
    },
    {
      "entropy": 5.906021928787231,
      "epoch": 1.019516072059343,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004902705105849517,
      "loss": 5.4707,
      "mean_token_accuracy": 0.1920158237218857,
      "num_tokens": 21526766.0,
      "step": 11545
    },
    {
      "entropy": 5.86635012626648,
      "epoch": 1.0199576121511833,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004902613480490553,
      "loss": 5.4835,
      "mean_token_accuracy": 0.20046430975198745,
      "num_tokens": 21535635.0,
      "step": 11550
    },
    {
      "entropy": 5.927623844146728,
      "epoch": 1.0203991522430236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902521812963116,
      "loss": 5.5415,
      "mean_token_accuracy": 0.19397743344306945,
      "num_tokens": 21544558.0,
      "step": 11555
    },
    {
      "entropy": 5.892992353439331,
      "epoch": 1.020840692334864,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004902430103268998,
      "loss": 5.6041,
      "mean_token_accuracy": 0.18775282502174379,
      "num_tokens": 21554539.0,
      "step": 11560
    },
    {
      "entropy": 5.951270961761475,
      "epoch": 1.0212822324267044,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004902338351410002,
      "loss": 5.5427,
      "mean_token_accuracy": 0.18847892731428145,
      "num_tokens": 21564305.0,
      "step": 11565
    },
    {
      "entropy": 5.772563791275024,
      "epoch": 1.0217237725185446,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490224655738792,
      "loss": 5.4222,
      "mean_token_accuracy": 0.20283106863498687,
      "num_tokens": 21573504.0,
      "step": 11570
    },
    {
      "entropy": 5.879359006881714,
      "epoch": 1.0221653126103851,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004902154721204552,
      "loss": 5.4966,
      "mean_token_accuracy": 0.19198623299598694,
      "num_tokens": 21582493.0,
      "step": 11575
    },
    {
      "entropy": 5.919539737701416,
      "epoch": 1.0226068527022254,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004902062842861696,
      "loss": 5.4659,
      "mean_token_accuracy": 0.19626332223415374,
      "num_tokens": 21591940.0,
      "step": 11580
    },
    {
      "entropy": 6.044213724136353,
      "epoch": 1.0230483927940657,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901970922361154,
      "loss": 5.5826,
      "mean_token_accuracy": 0.1809721127152443,
      "num_tokens": 21601747.0,
      "step": 11585
    },
    {
      "entropy": 5.919280099868774,
      "epoch": 1.023489932885906,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901878959704726,
      "loss": 5.5809,
      "mean_token_accuracy": 0.18843670636415483,
      "num_tokens": 21611477.0,
      "step": 11590
    },
    {
      "entropy": 5.983551502227783,
      "epoch": 1.0239314729777464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004901786954894212,
      "loss": 5.6603,
      "mean_token_accuracy": 0.17861845344305038,
      "num_tokens": 21622421.0,
      "step": 11595
    },
    {
      "entropy": 5.907370090484619,
      "epoch": 1.0243730130695867,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004901694907931416,
      "loss": 5.458,
      "mean_token_accuracy": 0.1950211003422737,
      "num_tokens": 21631677.0,
      "step": 11600
    },
    {
      "entropy": 5.881634283065796,
      "epoch": 1.024814553161427,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000490160281881814,
      "loss": 5.4451,
      "mean_token_accuracy": 0.18838044255971909,
      "num_tokens": 21641630.0,
      "step": 11605
    },
    {
      "entropy": 5.930839490890503,
      "epoch": 1.0252560932532675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004901510687556188,
      "loss": 5.6364,
      "mean_token_accuracy": 0.17747549265623092,
      "num_tokens": 21651176.0,
      "step": 11610
    },
    {
      "entropy": 5.9385966777801515,
      "epoch": 1.0256976333451078,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004901418514147367,
      "loss": 5.4841,
      "mean_token_accuracy": 0.2007811889052391,
      "num_tokens": 21659546.0,
      "step": 11615
    },
    {
      "entropy": 5.944651031494141,
      "epoch": 1.026139173436948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000490132629859348,
      "loss": 5.5432,
      "mean_token_accuracy": 0.18907264471054078,
      "num_tokens": 21668879.0,
      "step": 11620
    },
    {
      "entropy": 5.762916612625122,
      "epoch": 1.0265807135287883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004901234040896334,
      "loss": 5.4019,
      "mean_token_accuracy": 0.2002256453037262,
      "num_tokens": 21678259.0,
      "step": 11625
    },
    {
      "entropy": 5.861628818511963,
      "epoch": 1.0270222536206288,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901141741057737,
      "loss": 5.4783,
      "mean_token_accuracy": 0.19918524324893952,
      "num_tokens": 21688476.0,
      "step": 11630
    },
    {
      "entropy": 5.943405199050903,
      "epoch": 1.027463793712469,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004901049399079495,
      "loss": 5.5009,
      "mean_token_accuracy": 0.19515132308006286,
      "num_tokens": 21697936.0,
      "step": 11635
    },
    {
      "entropy": 5.919724369049073,
      "epoch": 1.0279053338043094,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004900957014963421,
      "loss": 5.5081,
      "mean_token_accuracy": 0.19047224074602126,
      "num_tokens": 21707171.0,
      "step": 11640
    },
    {
      "entropy": 5.857067203521728,
      "epoch": 1.0283468738961499,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004900864588711321,
      "loss": 5.4867,
      "mean_token_accuracy": 0.18926066756248475,
      "num_tokens": 21718341.0,
      "step": 11645
    },
    {
      "entropy": 5.88540678024292,
      "epoch": 1.0287884139879901,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004900772120325008,
      "loss": 5.4574,
      "mean_token_accuracy": 0.19791241139173507,
      "num_tokens": 21727104.0,
      "step": 11650
    },
    {
      "entropy": 5.890804767608643,
      "epoch": 1.0292299540798304,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004900679609806291,
      "loss": 5.5468,
      "mean_token_accuracy": 0.19061380624771118,
      "num_tokens": 21736933.0,
      "step": 11655
    },
    {
      "entropy": 5.87164273262024,
      "epoch": 1.0296714941716707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004900587057156985,
      "loss": 5.5017,
      "mean_token_accuracy": 0.18993761390447617,
      "num_tokens": 21746348.0,
      "step": 11660
    },
    {
      "entropy": 5.932830572128296,
      "epoch": 1.0301130342635112,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004900494462378902,
      "loss": 5.5645,
      "mean_token_accuracy": 0.1914984703063965,
      "num_tokens": 21755118.0,
      "step": 11665
    },
    {
      "entropy": 5.8500075340271,
      "epoch": 1.0305545743553515,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004900401825473854,
      "loss": 5.4756,
      "mean_token_accuracy": 0.1923319712281227,
      "num_tokens": 21764570.0,
      "step": 11670
    },
    {
      "entropy": 5.9224631786346436,
      "epoch": 1.0309961144471917,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900309146443658,
      "loss": 5.4667,
      "mean_token_accuracy": 0.19468381702899934,
      "num_tokens": 21772943.0,
      "step": 11675
    },
    {
      "entropy": 5.942181873321533,
      "epoch": 1.0314376545390322,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004900216425290128,
      "loss": 5.6388,
      "mean_token_accuracy": 0.18201075345277787,
      "num_tokens": 21782446.0,
      "step": 11680
    },
    {
      "entropy": 5.875915050506592,
      "epoch": 1.0318791946308725,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900123662015082,
      "loss": 5.43,
      "mean_token_accuracy": 0.19402869790792465,
      "num_tokens": 21791332.0,
      "step": 11685
    },
    {
      "entropy": 5.941959381103516,
      "epoch": 1.0323207347227128,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900030856620336,
      "loss": 5.5627,
      "mean_token_accuracy": 0.18612705320119857,
      "num_tokens": 21800776.0,
      "step": 11690
    },
    {
      "entropy": 5.877899837493897,
      "epoch": 1.032762274814553,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899938009107708,
      "loss": 5.3798,
      "mean_token_accuracy": 0.20429026782512666,
      "num_tokens": 21809035.0,
      "step": 11695
    },
    {
      "entropy": 5.857645177841187,
      "epoch": 1.0332038149063936,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899845119479019,
      "loss": 5.4844,
      "mean_token_accuracy": 0.18866659700870514,
      "num_tokens": 21819047.0,
      "step": 11700
    },
    {
      "entropy": 5.897699069976807,
      "epoch": 1.0336453549982338,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004899752187736085,
      "loss": 5.5918,
      "mean_token_accuracy": 0.19408478289842607,
      "num_tokens": 21828302.0,
      "step": 11705
    },
    {
      "entropy": 5.9029014110565186,
      "epoch": 1.0340868950900741,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004899659213880728,
      "loss": 5.5309,
      "mean_token_accuracy": 0.18749282956123353,
      "num_tokens": 21837371.0,
      "step": 11710
    },
    {
      "entropy": 5.84082612991333,
      "epoch": 1.0345284351819146,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004899566197914771,
      "loss": 5.4418,
      "mean_token_accuracy": 0.1880480945110321,
      "num_tokens": 21846234.0,
      "step": 11715
    },
    {
      "entropy": 5.949661922454834,
      "epoch": 1.034969975273755,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004899473139840033,
      "loss": 5.5268,
      "mean_token_accuracy": 0.18696542382240294,
      "num_tokens": 21856691.0,
      "step": 11720
    },
    {
      "entropy": 5.902682876586914,
      "epoch": 1.0354115153655952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004899380039658339,
      "loss": 5.4588,
      "mean_token_accuracy": 0.19653980135917665,
      "num_tokens": 21865587.0,
      "step": 11725
    },
    {
      "entropy": 5.870551824569702,
      "epoch": 1.0358530554574354,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004899286897371513,
      "loss": 5.4087,
      "mean_token_accuracy": 0.1995074048638344,
      "num_tokens": 21874644.0,
      "step": 11730
    },
    {
      "entropy": 5.950603818893432,
      "epoch": 1.036294595549276,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000489919371298138,
      "loss": 5.533,
      "mean_token_accuracy": 0.18489942252635955,
      "num_tokens": 21885149.0,
      "step": 11735
    },
    {
      "entropy": 5.842100620269775,
      "epoch": 1.0367361356411162,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004899100486489763,
      "loss": 5.4729,
      "mean_token_accuracy": 0.19274086654186248,
      "num_tokens": 21894048.0,
      "step": 11740
    },
    {
      "entropy": 5.777438068389893,
      "epoch": 1.0371776757329565,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004899007217898491,
      "loss": 5.3716,
      "mean_token_accuracy": 0.2060852736234665,
      "num_tokens": 21902941.0,
      "step": 11745
    },
    {
      "entropy": 5.8248457431793215,
      "epoch": 1.037619215824797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898913907209389,
      "loss": 5.4656,
      "mean_token_accuracy": 0.19611626118421555,
      "num_tokens": 21911160.0,
      "step": 11750
    },
    {
      "entropy": 5.91268630027771,
      "epoch": 1.0380607559166373,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004898820554424285,
      "loss": 5.4828,
      "mean_token_accuracy": 0.19892205893993378,
      "num_tokens": 21920151.0,
      "step": 11755
    },
    {
      "entropy": 5.861723756790161,
      "epoch": 1.0385022960084775,
      "grad_norm": 1.0,
      "learning_rate": 0.000489872715954501,
      "loss": 5.4873,
      "mean_token_accuracy": 0.1990682765841484,
      "num_tokens": 21930064.0,
      "step": 11760
    },
    {
      "entropy": 5.876028251647949,
      "epoch": 1.0389438361003178,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004898633722573391,
      "loss": 5.4899,
      "mean_token_accuracy": 0.1925680711865425,
      "num_tokens": 21938884.0,
      "step": 11765
    },
    {
      "entropy": 5.856072330474854,
      "epoch": 1.0393853761921583,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000489854024351126,
      "loss": 5.4855,
      "mean_token_accuracy": 0.19359561204910278,
      "num_tokens": 21948245.0,
      "step": 11770
    },
    {
      "entropy": 5.931502723693848,
      "epoch": 1.0398269162839986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004898446722360446,
      "loss": 5.5224,
      "mean_token_accuracy": 0.19026037603616713,
      "num_tokens": 21958788.0,
      "step": 11775
    },
    {
      "entropy": 5.909556150436401,
      "epoch": 1.0402684563758389,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004898353159122784,
      "loss": 5.5026,
      "mean_token_accuracy": 0.18899072408676149,
      "num_tokens": 21968192.0,
      "step": 11780
    },
    {
      "entropy": 5.873736190795898,
      "epoch": 1.0407099964676794,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004898259553800106,
      "loss": 5.442,
      "mean_token_accuracy": 0.20719268769025803,
      "num_tokens": 21978572.0,
      "step": 11785
    },
    {
      "entropy": 5.817206525802613,
      "epoch": 1.0411515365595196,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004898165906394244,
      "loss": 5.4215,
      "mean_token_accuracy": 0.20327729284763335,
      "num_tokens": 21987724.0,
      "step": 11790
    },
    {
      "entropy": 5.886700963973999,
      "epoch": 1.04159307665136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004898072216907033,
      "loss": 5.5081,
      "mean_token_accuracy": 0.19093688875436782,
      "num_tokens": 21997691.0,
      "step": 11795
    },
    {
      "entropy": 5.936474180221557,
      "epoch": 1.0420346167432002,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897978485340311,
      "loss": 5.5059,
      "mean_token_accuracy": 0.19006700068712234,
      "num_tokens": 22007568.0,
      "step": 11800
    },
    {
      "entropy": 5.900757646560669,
      "epoch": 1.0424761568350407,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000489788471169591,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18911885917186738,
      "num_tokens": 22017157.0,
      "step": 11805
    },
    {
      "entropy": 5.881508016586304,
      "epoch": 1.042917696926881,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004897790895975671,
      "loss": 5.4409,
      "mean_token_accuracy": 0.19429090321063996,
      "num_tokens": 22026059.0,
      "step": 11810
    },
    {
      "entropy": 5.84333028793335,
      "epoch": 1.0433592370187212,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004897697038181428,
      "loss": 5.4104,
      "mean_token_accuracy": 0.19652430415153505,
      "num_tokens": 22035425.0,
      "step": 11815
    },
    {
      "entropy": 5.885338497161865,
      "epoch": 1.0438007771105617,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004897603138315022,
      "loss": 5.458,
      "mean_token_accuracy": 0.19611585140228271,
      "num_tokens": 22044751.0,
      "step": 11820
    },
    {
      "entropy": 5.889741849899292,
      "epoch": 1.044242317202402,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004897509196378293,
      "loss": 5.5242,
      "mean_token_accuracy": 0.19303250163793564,
      "num_tokens": 22054449.0,
      "step": 11825
    },
    {
      "entropy": 5.935718822479248,
      "epoch": 1.0446838572942423,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489741521237308,
      "loss": 5.5317,
      "mean_token_accuracy": 0.18637033700942993,
      "num_tokens": 22064138.0,
      "step": 11830
    },
    {
      "entropy": 5.959399032592773,
      "epoch": 1.0451253973860826,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004897321186301223,
      "loss": 5.4993,
      "mean_token_accuracy": 0.19059466272592546,
      "num_tokens": 22073935.0,
      "step": 11835
    },
    {
      "entropy": 5.928448915481567,
      "epoch": 1.045566937477923,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004897227118164566,
      "loss": 5.5681,
      "mean_token_accuracy": 0.1886399656534195,
      "num_tokens": 22082774.0,
      "step": 11840
    },
    {
      "entropy": 5.942109632492065,
      "epoch": 1.0460084775697633,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000489713300796495,
      "loss": 5.5442,
      "mean_token_accuracy": 0.18486039489507675,
      "num_tokens": 22091799.0,
      "step": 11845
    },
    {
      "entropy": 5.9036217212677,
      "epoch": 1.0464500176616036,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897038855704221,
      "loss": 5.5211,
      "mean_token_accuracy": 0.19211660623550414,
      "num_tokens": 22100942.0,
      "step": 11850
    },
    {
      "entropy": 5.903837633132935,
      "epoch": 1.0468915577534441,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004896944661384222,
      "loss": 5.5701,
      "mean_token_accuracy": 0.1853426918387413,
      "num_tokens": 22110285.0,
      "step": 11855
    },
    {
      "entropy": 5.857434797286987,
      "epoch": 1.0473330978452844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896850425006797,
      "loss": 5.4953,
      "mean_token_accuracy": 0.19439102709293365,
      "num_tokens": 22119150.0,
      "step": 11860
    },
    {
      "entropy": 5.846440124511719,
      "epoch": 1.0477746379371247,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896756146573794,
      "loss": 5.4862,
      "mean_token_accuracy": 0.19287172108888626,
      "num_tokens": 22128290.0,
      "step": 11865
    },
    {
      "entropy": 5.836042785644532,
      "epoch": 1.048216178028965,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004896661826087059,
      "loss": 5.4375,
      "mean_token_accuracy": 0.1955834612250328,
      "num_tokens": 22138452.0,
      "step": 11870
    },
    {
      "entropy": 5.934262228012085,
      "epoch": 1.0486577181208054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896567463548439,
      "loss": 5.5102,
      "mean_token_accuracy": 0.19661574810743332,
      "num_tokens": 22148298.0,
      "step": 11875
    },
    {
      "entropy": 5.78337550163269,
      "epoch": 1.0490992582126457,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896473058959783,
      "loss": 5.2659,
      "mean_token_accuracy": 0.20811639875173568,
      "num_tokens": 22157448.0,
      "step": 11880
    },
    {
      "entropy": 5.863970756530762,
      "epoch": 1.049540798304486,
      "grad_norm": 1.0,
      "learning_rate": 0.0004896378612322942,
      "loss": 5.4837,
      "mean_token_accuracy": 0.2033226177096367,
      "num_tokens": 22167223.0,
      "step": 11885
    },
    {
      "entropy": 5.854445457458496,
      "epoch": 1.0499823383963265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004896284123639763,
      "loss": 5.4375,
      "mean_token_accuracy": 0.19668431282043458,
      "num_tokens": 22176312.0,
      "step": 11890
    },
    {
      "entropy": 5.853716278076172,
      "epoch": 1.0504238784881668,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048961895929121,
      "loss": 5.424,
      "mean_token_accuracy": 0.19919323474168776,
      "num_tokens": 22185805.0,
      "step": 11895
    },
    {
      "entropy": 5.826368713378907,
      "epoch": 1.050865418580007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004896095020141802,
      "loss": 5.3578,
      "mean_token_accuracy": 0.20796882212162018,
      "num_tokens": 22194763.0,
      "step": 11900
    },
    {
      "entropy": 5.86078109741211,
      "epoch": 1.0513069586718473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896000405330724,
      "loss": 5.4622,
      "mean_token_accuracy": 0.19625604450702666,
      "num_tokens": 22203114.0,
      "step": 11905
    },
    {
      "entropy": 5.887171411514283,
      "epoch": 1.0517484987636878,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000489590574848072,
      "loss": 5.5146,
      "mean_token_accuracy": 0.19038948714733123,
      "num_tokens": 22212641.0,
      "step": 11910
    },
    {
      "entropy": 5.894573116302491,
      "epoch": 1.052190038855528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000489581104959364,
      "loss": 5.4813,
      "mean_token_accuracy": 0.20056993663311004,
      "num_tokens": 22222107.0,
      "step": 11915
    },
    {
      "entropy": 6.0080546855926515,
      "epoch": 1.0526315789473684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895716308671343,
      "loss": 5.6151,
      "mean_token_accuracy": 0.1859358787536621,
      "num_tokens": 22231974.0,
      "step": 11920
    },
    {
      "entropy": 5.923385286331177,
      "epoch": 1.0530731190392089,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004895621525715685,
      "loss": 5.4635,
      "mean_token_accuracy": 0.19404913634061813,
      "num_tokens": 22241563.0,
      "step": 11925
    },
    {
      "entropy": 5.867195510864258,
      "epoch": 1.0535146591310491,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004895526700728521,
      "loss": 5.5077,
      "mean_token_accuracy": 0.19452970176935197,
      "num_tokens": 22250607.0,
      "step": 11930
    },
    {
      "entropy": 5.909785270690918,
      "epoch": 1.0539561992228894,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004895431833711708,
      "loss": 5.559,
      "mean_token_accuracy": 0.18584170043468476,
      "num_tokens": 22260528.0,
      "step": 11935
    },
    {
      "entropy": 5.8972796440124515,
      "epoch": 1.0543977393147297,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004895336924667107,
      "loss": 5.4145,
      "mean_token_accuracy": 0.1972587838768959,
      "num_tokens": 22269846.0,
      "step": 11940
    },
    {
      "entropy": 5.939662599563599,
      "epoch": 1.0548392794065702,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895241973596576,
      "loss": 5.5479,
      "mean_token_accuracy": 0.19393300712108613,
      "num_tokens": 22278520.0,
      "step": 11945
    },
    {
      "entropy": 5.927440214157104,
      "epoch": 1.0552808194984105,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895146980501973,
      "loss": 5.6062,
      "mean_token_accuracy": 0.18371869921684264,
      "num_tokens": 22287270.0,
      "step": 11950
    },
    {
      "entropy": 5.905961036682129,
      "epoch": 1.0557223595902507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895051945385163,
      "loss": 5.4277,
      "mean_token_accuracy": 0.19346169978380204,
      "num_tokens": 22296468.0,
      "step": 11955
    },
    {
      "entropy": 5.887745714187622,
      "epoch": 1.0561638996820912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004894956868248004,
      "loss": 5.5719,
      "mean_token_accuracy": 0.18321757465600969,
      "num_tokens": 22306323.0,
      "step": 11960
    },
    {
      "entropy": 5.901475143432617,
      "epoch": 1.0566054397739315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894861749092359,
      "loss": 5.4727,
      "mean_token_accuracy": 0.19942791759967804,
      "num_tokens": 22315377.0,
      "step": 11965
    },
    {
      "entropy": 5.917809677124024,
      "epoch": 1.0570469798657718,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004894766587920094,
      "loss": 5.4724,
      "mean_token_accuracy": 0.19685131460428237,
      "num_tokens": 22325328.0,
      "step": 11970
    },
    {
      "entropy": 5.868840837478638,
      "epoch": 1.057488519957612,
      "grad_norm": 0.96875,
      "learning_rate": 0.000489467138473307,
      "loss": 5.4928,
      "mean_token_accuracy": 0.20039464086294173,
      "num_tokens": 22334584.0,
      "step": 11975
    },
    {
      "entropy": 5.852491807937622,
      "epoch": 1.0579300600494526,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004894576139533154,
      "loss": 5.4452,
      "mean_token_accuracy": 0.20387588888406755,
      "num_tokens": 22343718.0,
      "step": 11980
    },
    {
      "entropy": 5.962774324417114,
      "epoch": 1.0583716001412928,
      "grad_norm": 1.015625,
      "learning_rate": 0.000489448085232221,
      "loss": 5.5936,
      "mean_token_accuracy": 0.1835495725274086,
      "num_tokens": 22352690.0,
      "step": 11985
    },
    {
      "entropy": 5.854324293136597,
      "epoch": 1.058813140233133,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004894385523102107,
      "loss": 5.3557,
      "mean_token_accuracy": 0.20705111622810363,
      "num_tokens": 22361736.0,
      "step": 11990
    },
    {
      "entropy": 5.815744400024414,
      "epoch": 1.0592546803249736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004894290151874711,
      "loss": 5.4531,
      "mean_token_accuracy": 0.1917500004172325,
      "num_tokens": 22370199.0,
      "step": 11995
    },
    {
      "entropy": 5.887319946289063,
      "epoch": 1.0596962204168139,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000489419473864189,
      "loss": 5.4831,
      "mean_token_accuracy": 0.19727725386619568,
      "num_tokens": 22380547.0,
      "step": 12000
    },
    {
      "epoch": 1.0596962204168139,
      "eval_entropy": 5.6932506435994785,
      "eval_loss": 5.5530900955200195,
      "eval_mean_token_accuracy": 0.19958059528962807,
      "eval_num_tokens": 22380547.0,
      "eval_runtime": 20.3388,
      "eval_samples_per_second": 1736.334,
      "eval_steps_per_second": 217.072,
      "step": 12000
    }
  ],
  "logging_steps": 5,
  "max_steps": 113230,
  "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": 4916842110443520.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}